1 7(6q Feb Iqq7 Sustainability Assessment: A Review of Values, Concepts, and Methodological Approaches Issues in Agriculture 10 BARBARA BECKER [tNSUI N AS 1 IVF AH GROUPI) ON INTIRNAI IONAI. AGRICUI FURA RI Sl.AR(A1l Aw Sustainability Assessment: A Review of Values, Concepts, and Approaches Issues in Agriculture 10 BRARARA BECKER About the CGIAR The Consultative Group on International Agricultural Research (CGIAR) is an informal association of 53 public and private sector members that supports a network of 16 international agricultural research centers. The Group was established in 1971. The World Bank, the Food and Agriculture Organization of tlhe United Nations (FAO), the United Nations Development Programme (UNDP), and the United Nations Environment Programme (UNEP) are cosponsors of the CGIAR. The Chairman of the Group is a senior official of the World Bank, which provides the CGIAR system with a Secretariat in Washington, DC. The CGIAR is assisted by a Technical Advisory Committee, with a Secretariat at FAO in Rome. The mission of the CGIAR is to contribute, through its research, to promoting sustainable agriculture for food security in the develop- ing countries. International centers supported by the CGIAR are part of a global agricultural research system. The CGIAR conducts strate- gic and applied research, with its products being international public goods, and focuses its research agenda on problem solving through interdisciplinary programs implemented by one or more of its inter- national centers in collaboration with a full range of partners. Such programs concentrate on increasing productivity, protecting the envi- ronment, saving biodiversity, improving policies, and contributing to strengthening agricultural research in developing countries. Food productivity in developing countries has increased through the combined efforts of CGIAR centers and their partners in devel- oping countries. The same efforts have helped to bring about a range of other benefits, such as reduced prices of food, better nutrition, more rational policies, and stronger institutions. CGIAR centers have trained more than 50,000 agricultural scientists from develop- ing countries over the past 25 years. Many of them form the nucleus of and provide leadership to national agricultural research systems in their own countries. Contents PAGE Introduction ..............................................1 Definitions and Concepts ..............................................2 Philosophical-Ethical and Sociocultural Considerations .................. ........6 Value of Nature .............................................6 Intergenerational Equity ................... ........................... I 0 Intragenerational Equity ............................................. 11 From Concept to Measurement ......................... ..................... 15 Context of System Theory .............................................. 1 5 Approaches to Sustainability Assessment ......................................... 20 Economic indicators .............................................. 21 Environmental indicators ............................................. 26 Social indicators ............................................. 29 Composite indicators and systems approaches ......................... 31 Ecosystem health ............................................. 34 Operationalizing Sustainability Assessment in Space and Time ..... 37 Criteria for Indicator Selection .............................................. 41 From Sustainability Assessment to Sustainability Policy ................ ........ 43 Ecology and Economy .............................................. 45 Short-Term and Long-Term Goals .............................................. 46 Local and Global Goals ................ ............................. 48 Conclusions .............................................. 49 References ............................................. 52 About the Author .............................................. 63 List of Tables and Figures Table 1. Indicators and Parameters for Sustainability Assessment ................ 22 Table 2. Criteria for the Selection of Sustainability Indicators ............... ...... 42 Figure 1. Conceptual Framework for Sustainability Assessment .................... 2 Figure 2. Normative and Scientific Aspects of Sustainability ..........................5 Figure 3. Space and Time Matrix for Sustainability Assessment ............. ..... 37 i Sustainability Assessment: A Review of Values, Concepts, and Methodological Approaches BARBARA BECKER Introduction Since sustainable development became the catchword in interna- tional discussions, several approaches to sustainability assessment have been developed. In order to measure or predict the sustainability of a land use system or a society, one must consider the inherent problem.s of ex ante analysis of complex systems. Appropriate scales and time horizons must be chosen; the preconditions and requirements for operationalization and quantification of sustainability must be defined; and the philosophy and value system behind this concept and its translation into policies must be made explicit. On the other hand, the ethical and political convictions behind the multitude of policy recommendations made under the umbrella of sustainable development often remain obscure. There is a need to develop criteria that can be used to indicate to what degree strategies and policies con- tribute to sustainable development. This paper helps clarify the conceptual background and the implications of the prevalent sustainability paradigm; and the termi- nology is analyzed to reveal underlying normative philosophical an(i political perceptions and intentions. To present the interdisciplinary nature of sustainability assessment, a conceptual framework (Figure 1) is proposed that is covered by the disciplines of ecology, econom- ics, and social sciences. All these disciplines are embedded in the poli- cy environment of a society and reflect its underlying ethical and cul- tural values. 1 Relating the different approaches to sustainability assessme..1t across disciplines and against the background of the conceptual frame- work allows us to appraise their relative potentials and limitations. A space and time matrix presents the scale and scope of the different methodologies used for sustainability assessment. The constraints to scientific operationalization of sustamnability and to its translation into policy measures, which are revealed by this reference system, highlight the necessity for continued integrated systems research. Figure 1. Conceptual Framnework for Sustainability Assessment r----------------------------------------I Ethical I cultural values I l l ~~~~~~~Sustainable l… development_ l l ~~~~~~~~~Socioeconomic l l ~~~~~~~~~~~well-being I i Environmental quality t l ~~~~~~~~~~~Policy environment X I I _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I L_______________________ _______ JI Definitions and Concepts The importance that the term sustatnabilitty has gained in lnter- national debate can be attributed to its use in the Brundtlanid Commission's report, Our Common Future (WCED 1987), which linked the term to development. This report emphasized rbe economn- ic aspects of sustainability by defining sustainable dlevelopment as "economic development that meets the needs of the present gener-a- tion without compromising the abilitv of future generations to meet their own needs." This combination of sustainability and develcp- 2 ment tries to reconcile economic growth in the neoclassical tradition with a new concern for environmental protection, recognizing tile biophysical "limits to growth" (Meadows et al. 1972) as a constraint to economic development. The term sustainability also was used in the CGIAR's mission statement in 1989 to mean "successful man- agement of resources for agriculture to satisfy changing human needs while maintaining or enhancing the quality of the environment and conserving natural resources" (TAC/CGIAR 1989). Earlier use of the term sustainability in ecological and agricultur- al literature had hardly been noted outside the scientific community directly involved. The term sustainability was used in the context of productivity, either as a descriptive feature of ecosystems, "sustain- ability is the ability of a system to maintain productivity in spite of a major disturbance (intensive stress)" (Conway 1983), or as "sustaina- ble yield" of agricultural crops (Plucknett and Smith 1986). These definitions have since been expanded to a comprehensive (yet hardly quantifiable) holistic concept (e.g., by the non-govern- mental organization [NGO] treaty [1993]) in an unpublished draft report "Agriculture is sustainable when it is ecologically sound, eco- nomically viable, socially just, culturally appropriate and based on a holistic scientific approach." Although this type of definition has been rejected as too vague by some scientists, it reflects the concern of many environmentalists and development agents to not separate society and environment, economy and ethics (Spendjian 1991). These three types of definitions represent the most common approaches; that is, economic, ecological, and holistic sustainability concepts, which are equivalent to the categories: Sustainable Growth, Agroecology, and Stewardship, as suggested by Harrington and oth- ers (1993) and Ruttan (1994). The concept of sustainability has its roots in forestry, fisheries, and range management. The most commonly agreed upon German equiv- alent term, Nachhaltigtkeit (though not identical in meaning and ety- mology), was first introduced in forestry by the miner von Carlowitz in the eighteenth century (Peters and Wiebecke 1983; Wiersum 1995; 3 BML 1995) to describe the maintenance of long-term productivity of timber plantations to continuously provide construction poles for the mining industry. This use of the term was driven by the same political interest in economic growth as the World Commission on Environment and Development (WCED) report 200 years later. The etymological roots of sustainability as a derivation from the Latin verb sustenere (= uphold) are discussed by Redclift (1994). This etymology is also reflected in the debate among Spanish-speaking sci- entists; that is, whether sostenibilidad (from sostener) or sustentabilidad (from sustentar) is the more accurate translation. The first term is closer to the passive connotation of "being upheld," while the latter reflects more the active aspect of"to uphold." These considerations of terminology indicate that there is a strong normative component in the concept of sustainable develop- nfent. This value-driven normative aspect makes sustainable develop- ment attractive for policymakers because it permits a direct transla- tion of political objectives into a broadly agreed upon overall con- cept. However, the normative approach has two severe disadvan- tages. First, it can be misused for ideological objectives and economic interests that are far from the original ideas of sustainability (e.g., an advertisement campaign of a chemical company). Second, the nor- mative aspect impedes an "objective" or "neutral" scientific analysis of the concept, which is the basic difficulty for scientifically sound sustainability assessment. Thus, a critical analysis of the normative concept of sustainability is required. Figure 2 shows the relationship between the normative and sci- entific aspects of sustainable development and the development of the terminology and definitions. On the normative side, two early political documents of the international environmental debate are cited as predecessors of the WCED report: (1) the Ramsar Convention of 1971 on the protection of wedands and (2) the docu- ments produced following the first United Nations (UN) conference on the environment in 1972 in Stockholm (cf. Wolters 1995). These documents spoke of wise use of natural resources and of environmen- 4 tally sound strategies, respectively, terms that were much more obvi- ously normative than sustainable development as the overall paradigrn for the second UN conference on the environment, held in Rio de Janeiro in 1992. Figure 2. Normative and Scientific Aspects of Sustainability 'Nachhaltigkeit' (von Carlowitz, 18th century) wise use (Ramsar Convention, 1971) sustainable yield (ecological definition) sound strategies d (Stockholm, 1972) sustainable development (WCED 1987) definition) (holistic definition) normative' 'scientific' vision X paradigm < operationalization ethical preconditions: indicators / parameters: -value of nature -environmental -intergenerational equity -economic -intragenerational equity -social policy new terminology ? 5 The combined term sustainable development was coined in the "World Conservation Strategy" of the IUCN in 1980 (Haber 1995), but it never gained paradigmatic appeal before its use and interpreta- tion in the WCED report. Since then, in addition to its political impact, the term rapidly became a new research paradigm in a wide range of disciplines, from the social sciences to biology (cf. Kuln 1962 and 1969; Norgaard 1989; Vedeld 1994). To be scientificaLly sound, however, the new paradigm must be operational. Because the term sustainability has recently become somewhat discredited by the obscuring ambiguity of normative and scientific aspects, there is a move to replace it. At first, it seems appealing to return to the term wise use for the normative component. However, this term has been usurped by a broad, conservative, anti-environ- mentalist movement in the United States (Brick 1995). Although a new term may be justified, there is still too little consensus on an alternative. Thus, for the time being, sustainability is still the most powerful concept for agricultural research and development, despite its limitations and the potential for misinterpretation. Based on the three representative definitions, three aspects will be discussed in order to translate normative concepts into scientific categories or, vice versa, to detect the ethical values and political con- cepts behind the (apparently) objective and neutral scientific assess- ment of sustainability. The first aspect when dealing with the value of the environment is the conceptualization of nature; the seconcL is the temporal dimension of inteTgenerational equity; and the thircL is the spatial or social aspect of intragenerational equity. These aspects relate to the scientific operationalization of sustainability from eco- logical, economic, and social points of view, respectively. Philosophical-Ethical and Sociocultural Considerations Value of Nature The current debate on the value of nature as a basic precondition of sustainability assessment can be characterized by two extreme posi- 6 tions. On one side the environment is regarded as a pool of resources that can be exploited by man for maximum economic prosperity, hence the demand for sustainability to maintain the environment and its productivity. On the other hand, nature is considered a value for its own sake, but is threatened by the increasing human population and the destruction and consumption of natural resources. Both positions need careful analysis to determine their impact on current approaches to sustainability assessment and policies. Hampicke (1993, 1994) revised the current literature on ecologi- cal ethics in view of the economic valuation of nature conservation. Five philosophical concepts are commonly distinguished: (1) theologi- cal arguments, (2) a pathocentric position (animal rights movement), (3) biocentric individualism, (4) biocentric holism, and (5) anthro- pocentric arguments (cf. Birnbacher 1989; Muller 1993). Different approaches to ecological ethics in the Western world have been developed during the last few decades, as religion lost its hitherto unquestioned uniting power as an ethical principle and as the ecological crisis became apparent (Fraser-Darling 1969). Although, fol- lowing Hampicke (1993) and Birnbacher (1980), in a liberal socieTy theological arguments can no longer be used as a commonly agreed upon basis for deriving secular laws on environmental protection, they still contribute important aspects to environmental ethics. In contrast to other approaches to ecological ethics, only the theological position regards nature as creation. Regarding nature as creation implies a cre- ator, and thus man's ultimate responsibility is toward the creator instead of toward other creatures as moral subjects (Birnbacher 1980; Hampicke 1993). With respect to sustainability, the belief in a creator implies that only this creator in his sovereignty can sustain his creation. Although this belief does not release man from a responsible behavior toward the creation, it relieves him from the unbearable burden of maintaining the life on earth. Important concepts in the discussion on sustainable development originated in theological ethics, such as the concept of stewaredship (Fraser-Darling 1969). Similarly, sufficiency concepts (as opposed to effi- 7 ciency concepts) as a normative approach for sustainable development have a strong affinity with religious world views (e.g., Sachs 1993). Nelson (1995) pointed out that efficiency, as the foremost economic paradigm of the modern world, has replaced religious principles. In such a neoclassical world view, according to Nelson, "the possibiliTy that consumption should be reduced because the act of consumption is not good for the soul, or is not what actually makes people happy, has no place within the economic value system." Of the other ecological ethics approaches, biocentric holism is the most relevant philosophy in the sustainability debate, as compared with the animal rights movement or with biocentric individualism. All these philosophies contrast with anthropocentrism in that they consid- er non-human beings as moral subjects with an intrinsic value. The principle of biocentric holism is summarized by Leopold (1949): "A thing is right if it trends to preserve the integrity, stability, and beauty of the biotic community. It is wrong if it trends othenvise."l Shearman (1990) discussed the dichotomy of anthropocentrism versus non-anthropocentrism applied to the sustainability debate. He concluded that non-anthropocentrism in the sense of Leopold is not a valid condition for sustainability because it is based on an intuitive appeal and lacks rational support. Furthermore, Hampicke (1993) showed clearly that-similarly to theological arguments-biocentric holism cannot be accepted as a common basis for today's liberal society because in its final consequence, it would lead to nondemocratic authoritarian policies for its implementation; in the extreme case it would lead to some form of ecofascism. However, biocentric holism is widely accepted by environmentalists as, for example, shown by Flitner (1995), who analyzed the contributions to the well-known book on In a drastic form, the principle of biocentric holism and its consequence was expressed by Nietzsche: "Es sind schon viele Tierarten verschwtunden; gesetzt daJl auch d'er Mensch verschwiinde, so wurde nichts in der Welt fehlen" (quoted by Birnbacher 1989, p. 404). Author's translation: "Many animal species have disap- peared in the past; taken the case that man disappeared, too, nothing in the world would be missing." 8 biodiversity edited by Wilson (1988), with the result that about a third are based on biocentric holism. Because neither theological arguments nor biocentric holism pro- vide a basis for consensus in today's society, anthropocentrism must be evaluated. This position has dominated occidental philosophy from its beginning (Miller 1993). In particular, Kant's "categoric imperative"2 is, in agreement with Hampicke (1993), the only common basis for democratic societies. According to Hampicke (1994), it is not neces- sary to decide if nature can be valued as a good in itself or as an instru- ment for the benefit of mankind in order to develop environmental policies. Human-centered arguments are sufficient as a point of depar- ture for action (cf. Turner and Pearce 1993; RSU 1994). The same conclusion-that anthropocentrism is sufficient and is the only common ground for the interpretation of the environment in the context of sustainable development-was clearly expressed in the first paragraph of the United Nations Conference on Environment and Development (UNCED) Rio Declaration: The CONFERENCE ON ENVIRONMENTAND DEVEL- OPMENT..proclaims that: Principle 1: Human beings are at the center of concern for sus- tainable development. They are entitled to a healthy and pro- ductive life in harmony with nature. This statement was strongly opposed by biocentric holistic environ- mentalist groups during the preparatory process. Although anthropocentric arguments are the only agreed upon basis for consensus in society, and although the duty toward humani- ty-as compared with the duty toward a creator or toward nature-is 2 "Act only on that maxim whereby thou canst at the same time will that it should become a universal law" (Kant 1785). 9 sufficient basis for policy development, there are still different positions being held with regard to the value of nature as a pool of resources (exploitable) as opposed to the recognition of immaterial values. This is reflected by the scope of monetarization of natural resources and amenities as a basis for economic sustainability assessment. The concept of limited natural resources (Meadows et al 1972) was taken up in economic theory, which recognized that scarcity or limited availability applies not only to human labor and capital but also to natural resources, including the sink capacity of the environ- ment (Daly 1991; Haber 1994). This recognition gave rise to new eco- nomic approaches by "ecological economists" (Costanza et al. 1991). However, this approach still is based on the value system of the neo- classical economic tradition; that is, it "rejects the idea that some things are literally priceless" (Nelson 1995). Although this economic perspective is central to the operatio- nalization of sustainability as a valuation of the environment for pre- sent and future use, there are still differences among economists (and ecologists) about how far and in what way the monetarization of na - ural resources is possible, meaningfuil, and legitimate. In the extreme case, this may lead to "knowing the price of everything but the value of nothing," as Oscar Wilde stated (quoted by Redclift 1994). This issue was raised recently in CGIAR discussions on water managemenL, when identifying the development of water accounting standards as a priority of the Inter-Center Initiative on Water Management (IIMI 1995). Intergenerational Equity The next ethical issue in the analysis of sustainability assessment, after valuation of the environment, is the demand for intergenerational equity, an entirely new aspect in international debate. This demand, as a duty toward humanity, goes beyond the traditionallv accepted kin- ship care for the next generation (cf. Heinen 1994). Although inte-- generational equity lies within the anthropocentric approach, its philo- sophical justification as a commonly agreed upon basis for society to derive laws and policies may be debated. 10 Two philosophical principles are used to justify the duty toward future generations. The first principle is again the categoric imperative by Kant. The second is the theory of justice by Rawls (1971) as an extrapolation of Kant's philosophy, which was not developed initially to address intergenerational justice (Hampicke 1994). Rawls presented a hypothetical case in which a group of people design a future society and distribute the available resources with the expectation that they themselves, in their second lives, will have to live in that society with- out knowing what their social position will be. He concluded from that scenario that such a society will provide, comparatively, the most favorable conditions for their least-well-off members. Although, at first glance, these two principles appear convincing as poticy principles for intergenerational equity, there are two severe shortcomings. First, their implementation cannot be forced by fear of revenge or outbreak of anarchy if the principle is not adhered to because the human beings of future generations cannot retaliate for any injustice done to them. Thus, these principles are more an appeal for moral duty than policy principles because they lack the element of egotism in the sense of Hume (Hampicke 1994). Second, Rawls's model does not take into account the dynamics of society, when in fzact social conditions and environmental goods change with time so that the extrapolation of current values to future generations may lead, in retrospect, to undesirable imbalances (Reddift 1994). Thus, both prin- ciples, in line with the normative "imperative of responsibility" Uonas 1984; cf. Christen 1996), serve as moral appeals to the present under- standing of "just" resource use, but they cannot be demanded as policy consensus from society if its members do not agree to such principles. Intragenerational Equity Intragenerational equity as an ethical demand is not a new issue specific to sustainable development; there are numerous publications on this topic in the literature of the social and political sciences. In the context of sustainable development, however, at least three issues need to be discussed: 1. Although the WCED report viewed intragenera- 11 tional and intergenerational equity as equally impor- tant, the spatial and social dimensions (intragenera- tional equity) tend to be neglected for the sake of the time dimension (intergenerational equity) (Dietz et a. 1992; Haber 1995; Hailu and Runge-Metzger 1993). 2. The sustainability debate currently is dominated by the industrial countries, on the normative-political side in the neoclassical tradition to ensure global eco- nomic development and on the scientific-theoretical side in the occidental tradition of value judgment and the perception of nature. In a non-western cul- ture, where each species is viewed as a spiritual being and future generations are regarded as spiritual con- temporaries, the value of nature and the moral responsibility toward future generations are consid- ered differently than in the occidental tradition. If, in that society, these views are commonly agreed upon, laws and social rules for sustainability policies may be based on this vision. However, given today's political conditions, such local sustainability policies are likely to remain minority exceptions. Redclift and others (1994) showed how local sustainability strategies are suppressed by national and international pressures and policies. Redclift also claimed that re-empower- ing local populations would result in their becoming stewards of their environment again rather than to remain poachers. These aspects relate to the underly- ing concept of science as pointed out by Feyerabend (1987), who makes the distinction between "scientif- ic" and "traditional" knowledge (Redclift 1994a). If intragenerational equity is taken seriously, the dis- tribution of power needs to be revealed and ques- tioned. Foucault (1981) showed the relationship between knowledge and power for modern sciences 12 (Redclift 1994a; Flitner 1995), which is clearly reflected by the sustainability debate. Sustainable development is defined on the basis of occidental sci- ence and value systems. Agarwal (1993) expressed the view that it is "a Western design to keep India backward" and that "in their desperate attempt to survive today, people are forced to forsake their tomorrow and overuse their environment." 3. The issue of power and participation is closely linked to the third relevant issue of intragenerational equity in the sustainability debate: unequal spatial endow- ment of natural resources, unequal access to these resources, and unequal profit gained from their use (Flitner 1995; Heinen 1994). Haber (1994) used the importation of minerals into the European Union (EU) as an example of the imbalance in resource dis- tribution and exploitation. A recent example of con- flicting interests in the use of local resources is the exploitation of petroleum in southern Nigeria, where the local population is denied sustainable manage- ment of their environment for the sake of energy provision, primarily to industrial nations, and for the profit of the national government. Unequal distribu- tion of and access to resources also applies at smaller scales; for example, between the rural and urban populations within a country and between men and women at the household level. These conditions must be taken into account when defining sustain- ability strategies. In economic terms this problem is translated into the relationship between local and global needs and strategies, or, as expressed by Costanza (1991), "making local and short-term goals consistent with global and long-term goals." Strategies for sustain- able development must fulfill this demand. 13 Returning to the valuation of nature, such a strate- gy implies a common value system, which to the economist is implicitly the monetary value of resources. However, monetary valuation mav not completely capture the true value of environmental goods, considering the appreciation of non-mone- tary values (e.g., to members of a reciprocal subsis- tence culture as opposed to participants in a market economy). In analogy to Hampicke's conclusion that an anthropogenic approach is sufficient to justify con- cern for the environment, it can be concluded that consideration of the interests of powerless contem- poraries and care for their needs indicates the validi- tV and credibility when claiming concern for future generations. If the rights of and participation by today's powerless are denied in sustainability strate- gies, then it must be asked if such strategies, suppos- edly for the benefit of future generations, are not guided by self-interest. At the same time a portion of self-interest has been shown as a valid common basis for policy develop- ment in democratic societies. Neither ethical appeals for altruism for contemporary human beings (and even less for future generations) nor controversial views of nature are valid bases for sustainabilitv strategies. Such policies must be based on (recipro- cal) individual and communal benefits, on costs and incentives, and on legislative measures. Any policy for sustainable development is subject to value judg- ment. Sustainabilitv assessment is not pure, neutral, objective science, but rather reflects these implicit value judgments. From these consider- ations, criteria can be derived for decisionmaking among the typical alternatives in sustainability strategies: 14 * Between the interests and rights of human beings and other species. * Between the interests and rights of present and future generations. * Between the interests and rights of different social groups. From Concept to Measurement In the preceding section the normative aspects of sustainabiliry and its assessment were pointed out. Only then should its scientific operationalization be considered. Scientific operationalization will be viewed in the context of system theory in order to adequately capture the complexity of sustainability in time and space, and its tradeoffs among different components and aggregation levels. Although it must be taken into account that even apparently holistic system analysis is reductionist in nature (cf. Roling 1994), it is considered the most appropriate "meta-language" (Lantermann 1996) for an interdiscipli- nary approach to sustainability assessment. Context of System Theory Conway (1983) used the term sustainability to describe a charac- teristic of an agroecosystem long before it was used in the context of political development. According to Conway (1985), sustainability is a measurable agroecosystem function in addition to productivity, sta- bility, and equitability. This definition and interpretation have been widely accepted for sustainability assessment of agricultural systems, and, therefore his approach will be analyzed in more detail (cf. Tisdell 1988). Conway uses the term in the sense of resilience, a systern's ability to respond to stress. Dalsgaard and others (1995) pointed out that Conway's interpretation of resilience is not consistent with its use in ecosystem theory: he relates resilience to maintaining pro- ductivity rather than to maintaining the structure and patterns of behavior (Holling 1987). Conway deliberately chose productivity trends as the most appropriate parameter for resilience because he 15 . suggests a pragmatic methodology for agroecosystem analysis in a participatory workshop setting. Similarly, the other system proper-- ties were selected to facilitate relatively quick and easy analyses oi^ agroecosystems, but not for sustainability assessment (Conway 1993). His examples of trend analysis use classical yield or price develop-- ment curves developed from measurements over the last decade or two (Conway 1985). These are typical ex post analyses with a short time frame, as he did not intend to discuss the long-term predictability oF system behavior. Therefore, his approach cannot immediately be applied to ex ante analysis with a long-term perspective. Ex ante analysis is the most difficult aspect of sustainability assess- ment. In terms of system theory, these difficulties relate to system properties such as complexity, interrelatedness, nontransparency, ancd dynamics due to feedback mechanisms, cumulative effects, time lags, or evolution (Dorner 1989). The predictability or extrapolation of system behavior is possible to only a limited extent. In principle, such predictability requires knowledge of the dynamics of the entire life cycle of the observed com- ponents, and needs to cover at least the time span of the life cycles. This is not always possible, however, in ex ante analysis of long-term developments. The most relevant experiences with respect to ecosys- tem management have been gained in fisheries, a typical field for the "tragedy of the commons" (i.e., the conflict between short-term indi- vidual interests and long-term common interests). Ludwig and others (1993) reviewed the concept of maximum sustained yield (MSY) for fisheries management based on their analysis of historical statistics. They concluded that this concept encouraged overexploitation of a fluctuating resource due to a "hatchet effect": the lack of limits on investment during good periods, but strong pressure not to disinvesL during poor periods, leading to a heavily subsidized industry that over- harvests the resource. They concluded that predictions of future events, in particular the effects of global warming and other possible atmos- pheric changes, are extremely difficult because the time scales involved 16 are so long that observational studies are unlikely to provide timely indications of required actions or the consequences of faling to take remedial actions. Wissel (1995) showed the potential of ecological modeling for predicting system behavior. In particular, he pointed out that mocLel- ing forces one to clearly define which system component or property is to be sustained and which time frame and spatial scale is to be used aLs a reference. While Conway (1985) stresses the need to sustain productiv- ity in agroecosystems, environmentalists often argue to sustain species number, composition, or spatial arrangement. In ecosystem theory it is generally agreed that systems may exist in any one of several stable states, depending on environmental con- ditions (multiple stability). As a consequence, small changes in fac- tors that influence the system may lead to sudden and strong changes or even collapse of the system. Such sudden changes have been expe- rienced in fisheries when a slight increase in the harvest rate has led to complete destruction of the fish resource (Wissel 1995). Thus, recognizing threshold levels of factors that may cause undesired or irreversible changes to an ecosystem is the greatest challenge in sus- tainability assessment. (This presumably, however, is not feasible.) In contrast to Conway's concept of resilience, in which produc- tivity declines slowly before breakdown (with different patterns of recovery), the challenge of sustainability assessment in agroecosys- tems is to detect hidden stress before it becomes apparent in yield decline or before the system is irreversibly damaged. Thus, for example, if we are to anticipate changes that are not yet apparent in yield decline, yield trend analysis must be complemented by indirect measures of the ecosystem's capacity to respond to stress (e.g., by observing symptoms of change not directly linked to the yield trend, such as species composition or rates of flow and turnover [Becker 1995; Dalsgaard et al. 1995]). Wissel (1995) concluded that modeling, under clearly defined conditions and assumptions, can be used to predict system behavior, 17 not as a deterministic prognosis but to indicate the level of probabil- ity of a system behaving in a particular way. This provides a way to assess the remaining risk of failure at applying certain strategies or policies. Such risk assessment has become common practice in environ- mental impact analysis over the last decade. However, predictability is limited not only by the risk of a more-or-less well known probabil- ity (e.g., the risk of drought in a region with long-term meteorologi- cal records), but also by uncertainty (i.e., unpredictable events whose effects cannot be assessed, either in quality or in intensity). An exam- ple of such uncertainty is the production and release of a new and potentiallv polluting chemical (Costanza 1993). How uncertainty can be operationalized in economic strategies was discussed by Funtowitz and Ravetz (1991) and Perrings (1991). Going beyond risk and uncertainty, Dovers (1995) included ignorance in designing sustainability policies. Ignorance refers to unprecedented effects that cannot be recognized (e.g., the toxic effects of DDT in the first years of its use before observing its harmful consequences on vertebrates [cf. Carson 1962]). Dorner (1989) showed the limitations of human cognition, in particular when dealing with the prediction of complex system behavior. Thus, there are inherent difficulties in ex ante analy- sis for sustainability assessment. Despite these shortcomings, system theory has proven valid for sustainability assessment. First, it contributes to clarifying the condi- tions of sustainability. By definition, system theory forces one to define the boundaries of the system under consideration and the hierarchy of aggregation levels. In agricultural land use systems the most relevant subsystems (or levels) are the cropping system (plot level); farming system (farm level); watershed/village (local level); and landscape/district (regional level). Higher levels (national, suprana- tional, and global) influence agriculture more indirectly by policy decisions or large-scale environmental changes (e.g., acid rain or global warming). Izac and Swift (1994) discussed the appropriate scale for sustainability assessment of agricultural systems in Sub- Saharan Africa. 18 By identifying the system hierarchy, externalities between levels and tradeoffs among components can be traced and explicitly taken into consideration. For example, in an agroecological system analyzed at the farm level, the effects of national policies are externalities as long as they are outside the decision context of the farmer (Olembo 1994). Typical tradeoffs among components within a farming system include unproductive fallow lands in a rotation system for the sake of soil recovery for future use. In resource economics the aspect of externalities has gained great importance in that methodologies are being developed to convert such externalities into accountable quan- tities (cf. Steger 1995), as well as the assignment of "opportunity costs" to tradeoff effects. Similarly, the "tragedy of the commons" (i.e., individual use of common resources) can be analyzed adequately only by considering the higher system level to find proper policies for sustainable use (e.g., the case of overgrazing in pastoral societies). Such conflicting interests among different groups-or hierarchical levels of the system-is a typ- ical problem in sustainability strategies. Problem analysis is greatly facilitated by system theory to derive alternative scenarios of future development, depending on the policy chosen. System analysis, in addition to indicating negative interactions of system components and conflicts among system levels, allows one to detect and describe synergies (i.e., positive interactions) among system components. In agroecosystems, crop mixtures with a Land Equivalent Ratio (LER), i.e. the total land area required in monoculture to produce the same yield as a given area of mixed crop, expressed as a ratio, greater than 1 would be an example of such synergetic effects. Ikerd (1993) pointed out that synergy is a crucial element in sustainable agriculture. To be used for sustain- ability assessment, however, synergy must be converted into mea- surable factors. These factors often are approximated by measures of the complexity of the system, which is considered a stabilizing quality. Although there is no immediate correlation between com- plexity and stability, the contribution of complexity and its syner- getic and stabilizing effects on system performance need to be 19 included in sustainability assessment (Dalsgaard et al. 1995; Piepho 1996). To demonstrate the difficulties of ex ante analysis, the deep-sea fishery industry was selected as an example because it is large scale and it has a long-term nature, thus it does not allow experiments with repli-- cations in space or in time. There are, however, systems with shorter cycles that can be described or modeled successfully by system analysis (e.g., cropping cycles in agriculture). In such cases, system theory can contribute to the predictability of system behavior, but it is necessary to choose the correct time scale for the purpose (cf. IBSRAM 1994). For example, although Izac and Swift (1994) mentioned the medium- term climatic cycle of 11 years in Southern Africa due to the ENSO (El Nifno - Southern Oscillation) effect of 10 to 15 years (Schonwiese 1995; Powell 1995), they arbitrarily proposed a 3-year basis of climatic differences for sustainability assessment. While Conway (1985, 1993) treated agricultural productivity as a dynarnic system property with regard to time, it should also be viewed as a spatially differentiated system property that reflects the resource endowment of the specific site. This includes not only the total amount of nutrients, but also the dynamics of biological activity and the genetic potential of crop varieties. Izac and Swift (1994), therefore, extended the concept of nondeclining productiviry in sustainable agroecosystems to byproducts and impacts on ecosvstem amenities, beyond harvestable outputs. Haber (1995) showed that intragenera- tional transfer (the spatially unbalanced resource endowment and unbalanced exploitation and transfer from the poor to the rich) is a potential reason for unsustainability. Approaches to Sustainability Assessment To present current approaches to the operationalization of sustain- ability, the-admittedly artificial but most common-division of eco- nomic, environmental, and social indicator concepts as well as com- posite indicators will be discussed. This is consistent with the underly- ing definitions; that is, focusing on economic, ecological, and social or holistic interpretations of sustainability. 20 Two basic approaches to sustainability assessment have bcen developed: * The exact measurement of single factors and their combination into meaningful parameters. * Indicators as an expression of complex situations, where an indicator is "a variable that compresses information concerning a relatively complex process, trend or state into a more readily under- standable form" (Harrington et al. 1993). The term sustainability indicator will be used here as a generic expression for quantitative or qualitative sustainability variables. The WCED (1987) definition, which focuses on intergenera- tional equity, and Conway's (1983) definition, which focuses on productivity trends, both concentrate on the dynamic aspect of Sl- tainability over time. Indicators to capture this aspect belong to the group of trend indicators, while state indicators reflect the condi- tion of the respective (eco)system (Bernstein 1992). In developing environmental indicators for national and international policies it has become common practice to distinguish pressure, state, and response indicators (OECD 1991; Adriaanse 1993; Hammond et al. 1995; Pieri et al. 1995; Winograd 1995). An overview on cur- rent sustainability indicators is presented in Table 1. Economic indicators. For economic trend indicators two basic approaches have been proposed: the valuation of discount rates of resource depletion and Total Factor Productivity (TFP). In the neoclassical economic tradition discount rates are derived from the concept of intergenerational equity, or more precisely, from its predecessor concept of limited unrenewable resources (Meadows et al. 1972). The concept of discount rates in the con- text of sustainable development was first proposed by Barbier (1989) and Pearce and others (1990). Assuming a certain "natural capital" stock of a given resource, rates of potential use are calcu- lated to maintain this resource for a given time before its deple- tion. Pollution rates also are calculated by assuming a given 21 Table 1. Indicators and Parameters for Sustainability Assessment ECONOMIC INDICATORS ENVIRONMENTAL INDICATORS * Modified gross national product * Yield trends * Discount rates * Coefficients for limited resources -depletion costs -depletion rates -pollution costs -pollution rates * Total factor productivity - Material and energy flows * Total social factor productivity and balances * Willingness to pay * Soil health * Contingent valuation method - Modeling * Hedonic price method -empirical * Travel cost approach -deterministic-analytical -deterministic-numerical * Bioindicators SOCIAL INDICATORS COMPOSITE INDICATORS * Equity coefficients * Unranked lists of indicators * Disposable family income * Scoring systems * Social costs - Integrated system properties Quantifiable parameters? * Participation * Tenure rights absorption or sink capacity of the environment and a desirable time horizon. In the last few years the aspect of sink capacity has gained increasing importance as a more serious environmental constraint than resource scarcity. The concept of discount rates for limited natural resources is based on the assumption that human economy is a subsystem of the finite natural global system (Daly 1991) "in which the limiting factor in development is no longer manmade capital but remaining natural capital" (Costanza et al. 1991). Most economists view manmade and natural capital as substitutes rather than as complements. If human 22 activity is viewed as a substitute for natural capital, then the scarcity of natural capital is weighted in a fundamentally different way. This has led to the distinction of "soft" and "hard" sustainability, the forrner assuming substitutability of human activity for natural capital, the [at- ter not assuming such substitution. Among economists the precise determination of discount rates has been an issue of the sustainabilitv debate in the last few years (e.g., Pearce 1990; Spendjian 1991; Steger 1995). This debate will not be repeated here, but it has clearly shown the different ethical values (i.e., the "weight" assigned to the demands of future generations and to the intergenerational transfer of [potential] wealth). Equally, the distinc- tion of "soft" and "hard" sustainability based on the assessment of the substitutability of human capital for natural resources reflects different concepts of the value of nature (van Pelt et al. 1995). If, for example, a rain forest species threatened by extinction is viewed as a potential source of a pharmaceutical product, it can be substituted for witd a synthetic production of this drug. If it is considered a living being with an intrinsic value, it can never be substituted for by human creativity and action. The concept of discount rates was developed in the context of global development and recognition of the scarcity of global resources (e.g., petroleum) or the absorption capacity of the atmosphere (e.g., for greenhouse effect gases). Thus, spatial scales are large and time frames are long (Figure 3). The second approach to economic trend assess- ment of sustainability, Total Factor Productivity (TFP), is at the other end of the space and time scale, at the farm level. It calculates the ratio of the total value of all outputs to the total value of all inputs for a given production system. The TFP approach was first suggested by Lynam and Herdt (1989). A modified version that included soil nutri- ents and land degradation in the valuation was presented by Ehui and Spencer (1990). The TFP approach has been criticized because it does not internalize external costs, such as environmental effects (Hailu and Runge-Metzger 1993). Herdt and Lynam (1992) tried to overcome this shortcoming by proposing the Total Social Factor Productivity (TSFP) as a more advanced approach than the TFP. The TSFP 23 included the environmental costs of production, but the question remains about how to value environmental costs appropriately anc. where to draw the boundary of internalization. Both approaches, TFP and TSFP, assume steadily increasing production, defining sustainability as the "capacity of a system to maintain output at a level approximately equal to or greater than its historical average" and "technology contributes to sustainability if ii: increases the slope of the trend line" (Lynam and Herdt 1989). From an ecosystem analysis point of view it is clear that no system with finite material resources can grow limitlessly without eventually collapsing. Thus, there is the underlying contradiction of sustainable growth in this concept. The same applies to Conway's (1993' upward productivity trend when he referred to resilience as sustain-- ability. Recently, a close correlation between the TSFP index and yield has been observed when the TSFP has been used to evaluate long-term agricultural experiments "because most of these experi-- ments have well controlled if not constant inputs. However, on typ- ical farms this would not necessarily be the case because inputs are constantly being adjusted" (Barnett et al. 1995). Still, yield trencL and TSFP refer to the same spatial and hierarchical level (i.e., the agroecosystem at the farm level) and cover similar time spans. If human economy is viewed as a subsystem of the global ecosys- tem where natural resources are considered material assets, then the question of adequate methods and values for their monetarization arises. Only after adequate monetarization can they be internaLized in the economic assessment of sustainable development. Differen. approaches have been suggested: modified gross national products, cost estimates for conservation and rehabilitation measures, and con-- tingent valuation methods. The Gross National Product (GNP) is the classical indicator or economic development; it counts environmental damages or social rehabilitation as positive values because they increase overall expenses of the national economy. This deficiency has been taken into accoun. in neoclassical economics by adjusting the GNP to create the Alodified 24 Gross National Product, subtracting resource depletion and pollution and considering income distribution (e.g., by an "index of sustainable economic welfare" [Costanza 1991]). This does not, however, solve the problem of accounting for nonmonetarized natural resources such as air or aesthetic values. Valuation of nonmonetarized resources is attempted either by cal- culating real or fictitious costs of production, conservation, and reha- bilitation of natural resources, or by assessing their subjective value to the population. The latter is carried out either by (1) surveys on the Willingness-to-Pay or Contingent Valuation Method for certain natural resources or amenities or (2) indirect evaluation of their appreciation (e.g., by the Hedonic Price Method or Travel Cost Approach [cf. Engelhardt and Waibel 19931). Costanza (1991) questioned the quali- tv of such results in that they do not adequately incorporate long-term goals because this methodological approach excludes future generations from claiming their interests. Engelhard and Waibel (1993) compiled a list of different approaches to monetarization of natural resources. They favored assigning monetary value to the entire natural environment as an instrument for policy planning and monitoring (e.g., they claim that it is more helpful to say that a spider has a value of US$10 than to assign it an "unappreciable" value). In contrast, under the assumption of "hard" sustainability, Hampicke (1994) concluded that each species is invaluable because of the risk of its extinction, and because it is impossible to predict the monetary value that future generations would assign to it if they chose to monetarize the value of species. Using biodiversity as an example, von Braun and Virchov (1995) showed the potential and limitations of monetarization and economic valuation of the different aspects of biodiversity. They presented a gra- dient of declining quantifiability, from the direct consumptive value of a species or its products to its potential genetic information for future generations. Economic sustainabiliry assessment is strongly value-laden. First, it assumes the possibility and right to monetarize all aspects of life and 25 nature (Nelson 1995). The conversion of sustainability aspects into monetary terms has the advantage that it permits comparison of and calculation with these different quantities in a uniform dimension. On the other hand, it assumes a congruent basis of calculation, which in reality differs considerably depending on the underlying value judg- ment. Furthermore, monetary values are not sufficiently consistent with ecosystem structure and function, and therefore their aggregation may lead to inadequate environmental policies (RSU 1994). Environmental indicators. Yield trend is the most obvious envi- ronmental indicator to assess the sustainability of agroecosysrems. However, its suitability must be questioned because yield trend can be assessed either by ex post analysis or by modeling. In both cases extrapolation of the results is risky because agricultural systems are not static and because environmental stress is not necessarily reflected by yield trend changes, and sudden collapse may occur (Wissel 1995). Yield trend is highly specific to the site and to the crop vari- ety, and thus is ultimately data intensive. Walsh (1991) and Hailu and Runge-Metzger (1993) pointed out that at least 20 to 25 years of observation are necessary to obtain results of 5 to 10 percent accura- cy. Barnett and others (1995) showed the potential and limitations of productivity trends for sustainability assessment under controlled environmental conditions by evaluating long-term agricultural exper- iments around the world. The calculation of discount rates of resouirce depletion and pollution can be used as an environmental, as well as economic, trend indicator. In this case, the dimension would not be monetary values but physical units (e.g., tons or parts per million). In contrast to yield trends with small spatial scales and short time spans, this approach is applied mainly for large-scale resource exploitation and (nonpoint) long-term pollution, such as gaseous emissions or global warming. Frequently, these physical calculations are used as a basis for economic valuation, in particular to extrapolate the potential and limitations of industrial development. According to Der Rat von Sachversrtndigen fur Umweltfragen (RSU 1994), physical indicators should be prioritized; monetary indicators should be used as complementary. 26 The search for environmental indicators originated with indus- trialization and the accompanying pollution. In the pre-indusrtial era, canaries were used in mines as an earlv warning system to detect increases in carbon monoxide so that miners could be evacu- ated. Canaries belong to the group of monofactorial bioindicators, the first generation of environmental indicators. This type of bioindicator includes species that react sensitively to changes in their environment (e.g., lichens to increased SO2 or heavy metals, or tobacco to increased ozone in the atmosphere). With increasing environmental awareness and experience the use of ecological indi- cators was further refined. The second generation of ecological indicators focused on ecosystem dynamics, on the structure and func- tion of entire ecosystems. Parameters for stress/response assessment were developed, and chemical compounds and processes or meta- bolic products were measured and quantified. This ecosystems approach included the assessment of values such as "purity of nature," "amenities," or "ecosystem integrity," as expressed by the "Index of Biotic Integrity" (Regier 1992). The latter ecosystem properties have strong normative components that require clearlv defined value systems. A comprehensive review of the different approaches to ecological indication is given in McKenzie and others (1992). Only recently has identification of ecological indicators been dis- cussed in the context of socioeconomic and cultural conditions (Rapport 1992); that is, the growing environmental movements in industrial nations "discovered" sustainability as a concept for environ- mental quality assessment. Thus, the concept of ecological indicators merges with the concept of sustainability indicators (which have come into the debate in the context of international development) into one coherent concept of global relevance (Becker 1995). In this sense environmental indicators for policy planning and monitoring comprise ecological indicators of pollution (first genera- tion), of ecosystem structure and function (second generation), and of socioeconomic aspects (third generation). Thus ecological indicators in the narrower sense are ecosystem descriptors, while environmental indi- 27 cators have a more general meaning, are policy-oriented, and are not necessarily based on ecosystem analysis. Recognizing the need for such policy-oriented indicators at the national and international levels, the Organisation for Economic Co- operation and Development (OECD 1991) compiled a list of some 20 indicators, defined as pressure/state/response indicators. For example, on an expost data basis, the OECD proposed emission rates of key gases to assess atmospheric pollution, nitrate concentrations for river quality assessment, and numbers of threatened species. These biophysical aspects are complemented by socioeconomic indi- cators, such as population data or energy intensity and supply. These indicators are clearly rooted in the tradition of the assessment of environmental damages by industrialization, oriented toward nega- tive tradeoffs of past and present developments. They are not guided by an objective or target future sustainability scenario. The OECD list is driven by data availability rather than being designed on a solid theoretical basis (RSU 1994). Target-oriented approaches to environmental sustainability assessment for national policymaking have been developed in the Netherlands. Although among the most advanced approaches to sustainability assessment for actual policy planning, they are based on static concepts and do not precisely reflect ecosystem behavior ("quick and dirty indicators," Kuik and Verbruggen 1991). Material and energy flows are important ecosystem properties and have been recognized for their relevance to sustainability assessment and policy development (e.g., Daly 1991). Henseling and Schwanhold (1995) proposed material flow management as a strategy for sustainable development. Although it appears directly derived from ecosystem theory, it is a pragmatic approach related to industrial production and trade, similar to Production Chain Analysis, as a moni- toring instrument for material flows (Meissner 1993; RSU 1994). In contrast to these not precisely ecological environmental indi- cators, the RSU (1994) claimed that it was necessary to orient indica- 28 tor selection on ecosystem theory. The RSU proposed indicators to reflect critical levels, critical loads, and critical structural changes at different levels of the ecosystem hierarchy (e.g., the frequency of ozone levels beyond a defined threshold, pesticide concentrations in rivers, or habitat integrity). They admitted, however, that the opera- tionalization of such indicators still requires considerable research, and therefore they proposed to use the most sensitive components of an ecosystem as stress indicators. Identifying stress symptoms in the most vulnerable and sensitive ecosystem element presents an early warning for the entire ecosystem. Thus, these indicators are represen- tative of the complete cause/response chain in that particular ecosys- tem. Becker (1995) showed how indicator plants may be used to detect agroecosystem change in a way that complements analysis of productivity trends. Social indicators. Social indicators are intended to translate aspects of intragenerational equity into measurable quantities, or at least into operationalized terms. However, approaches to quantifica- tion and operationalization of social dimensions must be carefully restricted to those aspects that can be described meaningfully by numerical or analytical tools and methods (Hardin 1991). The most direct quantification of equity involves calculation of the distribution of wealth within a society. Two similar parameters have been proposed: the less common Herfindahl Measure for absolute concentration and the Gini coefficient for relative concentration, derived from the Lorenz curve (Conway 1993; Muller 1995). The Gini coefficient tends toward zero in a perfectly egalitarian society and toward one at increasing inequality (Izac and Swift 1994). Winograd (1995) used the Gini coefficient to list the agricultural land concen- tration in Latin American countries and to show how it has changed over the last three decades. At first, such a numerical expression seems convincing as an indi- cator of equity. However, it has several shortcomings that must be con- sidered. From a mathematical point of view, the Gini coefficient tends to give relatively greater weight to changes in different parts of the 29 range (Conway 1993). More important, it is based on a static percep- tion of social and cultural values and conditions, and pretends total uniformity of people, which is obviously not valid. People differ in their endowment and in the way they use and appreciate their resources (cf. Nelson 1995), and they are conscious of social justice. Conway (1993) therefore preferred Atkinson's weighted index oi income distribution. Huber (1995) made the distinction of social jus-- tice based on need, on performance, and on property as different dimensions of equity. These differences are not taken into account in static, target-oriented sustainability policies (e.g., Levi 1995). In early economic theory, as conceived by Pareto (1935), eco-- nomic optimality included the maximization of economic efficiency along with the maximization of social welfare (Izac and Swift 1994), while in neoclassical economics the focus was more on economic effi-- ciency only (Nelson 1995), separating it from social welfare. Taking into account that these two aspects are different sides of one coin, eco- nomic efficiency measures have been proposed as social indicators as well. Two common approaches are disposable family income at the household level and the calculation of social costs (Hailu and Runge- Metzger 1993). Apart from the level of wealth (or poverty) and its dis-- tribution across the society, two other aspects are particularly relevanr. for social sustainability assessment: legal aspects of land tenure ancL participation. Although land tenure can be expressed by the Gini coef- ficient, its assessment in different societies with varying cultural ancd legal practices is a complex issue that cannot easily be operationalized. From a social science point of view, participation is a central ele-- ment of sustainability (Busch-Liity 1995). Participation, however, is difficult to translate meaningfully into quantitative terms as a social indicator. In particular, with respect to sustainable development, Led (1991) pointed out that participation does not ensure equity irn resource use, which is fundamentally tied to the land reform issue; thar is, tied to land tenure (Adger and Grohs 1994). Participation has also been demanded as an instrument o.f Environmental Impact Assessment (EIA). Since ex ante analysis oF 30 potential impacts of planned projects on the environment is difficult, participation is intended to reduce uncertainty by intrasubjective judg- ment; furthermore, participation increases the transparency of the deci- sionmaking process (Meissner 1993). Recent experience with partici- patory EIA of release trials of genetically modified crops has revealed, however, that the power structures in this process and the subsequent decisionmaking could not be overcome by integrating NGO groups into the process (Stoeppler-Zimmer 1993). Social indicators refer to all levels of the spatial hierarchy, with varying degrees of relevance. Equity can be assessed within a family or across a village community as well as among different countries. The temporal dimension of these social indicators is of comparatively minor importance. Composite indicators and systems approaches. None of the sin- gle indicators presented above can adequately represent the level of sustainability of an ecosystem or society. Therefore, several ways of combining or aggregating indicators have been proposed. Three basic approaches can be distinguished: unranked lists of heterogeneous indicators, scoring systems with unified dimensions, and system descriptors. The first, and simplest, approach to sustainability assessment by composite indicators are lists of indicators without the intent to aggre- gate or unify them into a single dimension, and without assigning dif- ferent weights to the different components. The common lists of envi- ronmental indicators belong to this group, as proposed by, for exam- ple, the OECD, the World Resources Institute (WRI), the United Nations Development Programme (UNDP), the World Bank, or the Food and Agriculture Organization of the United Nations (FAO) (Muller 1995). Such lists of indicators have been used in recent attempts to apply the information gained from the scientific sustain- ability discussion to policy development (e.g., RSU 1994; Winograd 1995; Bund and Misereor 1996). The advantage of these lists is their transparency. Because they are predominantly action oriented for prac- tical policy, they are based on available data and on the normative 31 principles of the generating institution, while the scientific operational- ization of sustainability is less obvious in the choice of components. The definition of threshold values of resource exploitation or pollution is often interest driven rather than based on "hard" scientific knowl- edge of resource availability or scarcity (e.g., the target to reduce petro- leum consumption of cars to three liters per hundred kilometers). Scoring system approaches combine different components of the "sustainability complex" into one measure; the components may be given different weights according to the objectives or preferences of the authors. Examples of scoring systems at the local level are the Sustainable Livelihood Security Index (SLSI) defined by Swaminathan (1991) and the Agroecosystems Analysis Framework (e.g., Tabora 1991). The SLSI combines ecological, economic, employment, and equity factors by weighting three components: car- rying capacity (human and animal), number of economically active adults in the village, and the degree of female literacy and employ- ment. The SLSI thus is a subjective measure of sustainability, or, more precisely, a measure of "sustainable food and nutrition security at the household level," as defined by the author. The Agroecosystems Analysis Framework is a scoring system that distinguishes biophysical aspects, economic and social impacts, and policy and administrative aspects, subdividing each category into four or five components with a certain score as a basis for comparison of different land use systems. This method, although more detailed than the SLSI, also is subjective in its choice of components and weights of scores. In the Netherlands, national-level scoring systems have been proposed (cf. van Pelt et at 1995) in which indicators similar to those in the unranked lists are calibrated to uniform dimensions by target values (e.g., in the AMOEBA approach [ten Brink 1991]). The problem with scoring systems is that they pretend objectivity and uniformity, while the choice of components and their assigned weights is highly subjective, and the aggregation of different spatial, temporal, and sectoral dimen- sions is often not meaningful (cf. Muller 1995). System-based approaches apply strict rules of system theory to select a number of system properties as sustainability indicators, and a 32 set of rules that specify how to integrate them into a meaningful assessment of the system's sustainability. Examples of this approach, which are described below, include the quantification of ecological sustainability in farming systems analysis by Dalsgaard and others (1995) and the system-based operationalization of economic indica- tors at different spatial scales by van Pelt and others (1995). Dalsgaard and others (1995) selected four system properties that they considered crucial for sustainability-diversity, cycling, stability, and capacity-and they explicitly explained their selection criteria based on ecosystem theory. Their methodology was applied in a par- ticipatory process with local farmers for sustainability assessment of local agroecosystems. They based their calculations on parameters that can be easily measured by the farmers themselves, but they used sophisticated mathematical models to convert the results into sustain- ability measures. Although the underlying assumptions used as the basis for selecting the four system properties could be debated (e.g., high species diversity is not necessarily correlated with sustainabiliuy, as indicated by stable, but species-poor, natural ecosystems), the authors' explicit ecosystem-based approach of critical system proper- ties is a promising step toward sustainability assessment. This method, however, is currently restricted to the local level, and does not explicit- ly consider economic or social aspects. Beyond that, it is based on the assumption that the chosen system properties of agricultural systems follow the same patterns as in natural ecosystems. Nor did the authors explicitly assess human intervention in farming systems. The strength of this approach lies in its systematic selection of system properties, and less in the applied set of rules for their integration. Its focus is on state indicators of system conditions, without considering trends over time. It is intended for spatial system comparison rather than for aria- lyzing the dynamics of the resource base in one site. Van Pelt and others (1995) used discount rates as basic parame- ters, considering the substitutability of natural resources by distinguish- ing between "hard" and "soft" sustainability. Identifying the normative decisions in this distinction (intergenerational equity) as well as the conditions of social welfare (intragenerational equity), they developed 33 an algorithm to derive policy decisions based on an integrated model of intra- and intergenerational equity for different spatial levels in a hierarchical order (multicriteria analysis [MCA]). They discussed the relative advantages and disadvantages of ecological sustainability indi- cators proposed by other authors. The strength of this approach lies more in the set of rules used to integrate the different aspects of sus- tainability into a coherent system than in the selection of sustainability indicators. The MCA model does not, however, consider the interac- tion between different system components. The model is clearly based on trend analysis with less emphasis on detailed description of system conditions. In both examples the final results are expressed as a single figure (i.e., in the same way as in scoring systems). The difference between the two approaches lies in the system-based selection and integration of the single components. As for the other scoring techniques, final fig- ures are most easily interpreted by the authors themselves, or they may be used as a relative measure for system comparison, as suggested in both examples. Ecosystem health. Even before sustainability became fashionable, environmentalists had proposed "ecosystem health" as a concept for managing environmental resources. Since then, the usefulness of this approach to assess environmental quality has been subject to debate (Shrader-Frechette 1994). There are strong arguments that favor such an approach; the most obvious is its intuitive appeal, which makes ecosystem health particu- larly attractive for policy recommendations. Proponents of this concept (e.g., Rapport 1992; Costanza 1991) stress its holistic perspective based on a positive vision instead of focusing on single degradation symp- toms. Rapport (1992) compared the three generations of ecological indi- cators to different stages of diagnostic experience in human health. According to his interpretation the first generation, monofactorial indi- cators, focused on "clinical signs" of environmental degradation. The 34 second generation, ecosystem structure and function assessment, detected "preclinical signs and symptoms" of ecosystem breakdown. The third generation, which sought to establish connectivity between ecological considerations and economic and social factors, tried to define a larger and more appropriate context for assessing the health of the environment. He proposed three requirements for future ecological indicators: early warning by single monofactorial indicator species and forecasting models, groups of indicator species to reflect multiple stress, and indicators for ecosystem health based on properties of mature ecosystems, such as increasing complexity, feedback, and so on (cf. Becker 1995). Despite the advantages of intuitive appeal, holistic perspective, and positive vision, the concept of ecosystem health must be applied with care in sustainability assessment. The term ecosystem health, like the term sustainability, is a strongly value-laden and normative con- cept. As human health may be defined differently by those with dif- ferent perspectives and by different interest groups (e.g., the smokers' lobby or a health insurance company), the meaning of ecosystem health is neither precisely defined nor unanimously accepted by the scientifc commuity. By origin, the ecosystem health concept has a strong affinity to Leopold's holistic biocentrism. Norton (1991) spoke of "metaphysical holism" when ecosystem health is defined in the sense of biocentrism, and he distinguished it from "contextual holism" as an ecosystem health approach based on system theory. Its use in the latter context is based primarily on resilience as the most relevant system property. Wolman (quoted by Shrader-Frechette 1994) suggested that resilience be complemented by categories such as persistence, equity, attention to scale, and compassion, but this combination of categories leaves the domain of system theory, in particular by introducing compassion as an ethical aspect. Operationalization of ecosystem health presents the same problems as the operationalization of sustainability. Compared with the definition of human health as a basis for medical intervention, the definition of ecosystem health for environmental management is even less clear. In contrast to human health, which is centered on the 35 individual patient as the target "system," the boundary and hierarchica[ level of the "patient" ecosystem is less obvious and there is less agree-- ment on it. Therefore, Norton (1991) pointed out that it is necessary to define the proper scale for ecosystem health using a system theory approach ("contextual holism"). Allen (quoted by Shrader-Frechette 1994) emphasized that it is necessary to explicitly define the spatial and temporal boundaries of the respective ecosystem. Thus, all that has been discussed above that pertains to sustain- ability assessment also applies to the concept of operationalizing ecosystem health. Barnett and others (1995) therefore concluded that "good indicators of ecosystem health are good indicators of sustainabil- iry." They proposed soil properties such as erosion, organic matter con- tent, pH, phosphorus, potassium, micronutrient availability, microbial respiration rate, and bulk densiry as ecosystem health indicators to assess the quality of the resource base. They pointed out, however, that such indicators, which are based on experience, do not permit precise conclusions and that a "soil health index" would require further study before validation. In this sense, the ecosystem health approach is valuable for sus- tainability assessment because it complements "hard" system analysis with "soft" heuristic data and indicators, which are derived from expe- rience and observation of symptoms. Somerville (quoted by Shrader- Frechette 1994) pointed out that ecology is not only science but also art, and that therefore ecosystem health is an appropriate approach to reflect the art aspect of environmental management. The combined emphasis on art and science is taken for granted in medical diagnosis and treatment; equally, ethical aspects have always been an integral part of medical science. Thus, sustainability assessment carried out using the ecosystem health concept may benefit from the experience and knowledge acquired in the political debate on human health. The controversial debate on the consideration of "alternative" medical approaches, on the economic value of human life, organs, and medical services, as well as on technical, mechanistic versus holistic medical treatment may be viewed as a source of insight for the debate on ecosystem health and sustainability. 36 Figure 3. Space and Time Matrix for Sustainability Assessment log yrs 4- 3-. hne C 22 ci~ ~ ~~~~~~gi 7i 2t ,e *a' hat, rra .E decades /. agric - StD° a a 0atimlyyo0 ci policy O yeears / raoiorrar trend deviation .E OjoOd-ida5.rs 0- -21 -3 -2 -1 0 1 2 3 4 log km 1m a10m loom 1km 10 km *00 km 1000km 10000km plot farm village td00%. g natio.a gobal E expenments lield research social science policy researcrh ocarn* research ) modeling r GIS / RS / modvilivg C .E tanners to E public policy 0 V community municipality national o t.. onal industry 0. pollution control resourCe supply Operationalizing Sustainability Assessment in Space and Time Sustainability indicators can be divided roughly into state and trend indicators and have specific spatial scales and time horizons. Figure 3 presents current sustainability indicators in a space/time frame; a logarithmic scale was chosen so that the entire range of local to global dimensions and of short to long durations can be shown. 37 Although short time spans (i.e., hours or less) are relevant for biological and physical processes (e.g., soil biology or sudden point-source pollu- tion), sustainability assessment is focused on longer time periods. Thus, these fast processes are not covered by this diagram, although they may enter into the modeling of system components. The dashed lines indicate agriculture and policy, two realms of human activity that are central to sustainability research. Agriculture comprises spatial scales from plot-level and small-scale gathering to agribusiness and large plantations, and temporal scales of short-season crops to perenni- als and long rotations in swidden agriculture (cf. Fresco and Kroonenberg 1992; Fresco 1994). The diagram shows that the two approaches central to the assess- ment of agroecosystem sustainability-yield trend deviation and Total Factor Productivity-are very limited in scale and scope. I i particular, they cover time spans of only one or two decades. Thus, they do not assess intergenerational change, which, according to the WCED definition, is the principal criterion for sustainability. Similarly, policy measures have a temporal scope of at most one to two decades, but generally much less (e.g., one legislative period). The most prominent approach derived from the concept of intergen- erational equity is to use discount rates of resource depletion cr absorption capacities for pollution. This concept can be applied to trend assessment of abiotic, physical, and terrestrial resources and to the assumed sink capacity of aquatic, terrestrial, and atmospheric sys- tems over intergenerational timne periods. Only recently has the importance of biodiversity for sustain- ability been recognized. The focus on the Biodiversity Convention_ at the United Nations Conference on Environment and Development (UNCED) can be viewed as a reflection of this new awareness. This strong interest in biodiversity as a crucial elemert for sustainable development was generated at UNCED mainlv by the industrial countries, not only for conservationist reasons but also, and primarily, for economic reasons; that is, to maintain genetic resources for future use and exploitation. Redclift (1994) pointed out that the UNCED follow-up agenda of the Global 38 Environmental Facility (GEF) is dominated by the interests of industrial countries, one of which is to halt genetic erosion. Flitner (1995) documented how the history of the conservation of plant genetic resources, since their "discovery" a century ago, has been driven by economic interests. Although research on biodiversity has been high on the scientific agenda for the past decade (e.g., Wilson 1988; Schulze and Mooney 1993; Gaston 1996), the link between biodiversity and sustainability assessment is still weak. Biodiversity has an ambiguous role in sustain- ability assessment. On the one hand, it is in the focus of what needs to be sustained. On the other hand, biodiversity is proposed as a means to assess the sustainability of complex systems. Considering the threat of irreversible species extinction, the con- servation of biodiversity is the greatest challenge for sustainability strategies with the longest-term implications. For sustainability assess- ment focusing on biodiversity, the Center for International Forestry Research (CIFOR), inter alia, suggest to distinguish threat assessment and sensitivity analysis: the (static) assessment of threat to genetic resources in situ and the (dynamic) analysis of the impact of changing conditions (e.g., land prices, access to markets, etc.) on the level of threat (Boyle 1995). With regard to the use of biodiversity as a parameter in sustain- ability assessment, several approaches were presented above. Single species are used as monofactorial pollution indicators; groups of species are used to reflect multiple ecosystem stress. Applied to plant commu- nities, such observations may include phytosociological, structural parameters that go bevond simple counting of species numbers (Becker 1995). To assess sustainability at the local level it may be necessary to analyze biodiversiry below the species level (i.e., intraspecific diversity). Patterns of phenotypical or genotypical differences and their relation to site conditions may be indicative of forces that determine if a systenm is sustainable. Diversity at the variety level of crops can indicate sustain- able land use, reflecting ecological site conditions as well as the socioe- conomic conditions of farmers (de Boef et al. 1993). 39 Species numbers are the basis of biodiversitv for sustainability assessment as proposed by Daalsgard and others (1995), OECD (1993), and Winograd (1995). Daalsgard and others (1995) use the Shannon Index, while Piepho (1996) pointed out that this combined index may be a misleading parameter for biodiversity in agricultural systems. OECD (1991) used the percentage of threatened species in relation to all species of a country, admitting, however, that these data are not standardized internationally. Winograd (1995) complemented the number and ratio of threatened and rare species per country with the Species Risk Index, which is defined as the number of endemic species per unit area multiplied by the percentage of loss of original area (Reid eta!. 1992). Loss of habitats by conversion into agricultural land or by intensi- fication of production is the major cause for extinction hazard (OECD 1991). The RSU (1994) therefore concluded that the assessment of structural landscape change should be a crucial element in the opera- tionalization of sustainability. The decision about which landscape structure is to be sustained is highly culture dependent and value-dri- ven. Apart from regions with extremely adverse conditions for human settlement, landscape patterns have been created by mankind over mil- lennia. These regions have undergone "creative destruction" (cf. Schumpeter quoted by Fritz et al. 1995), which in many cases has increased their biodiversitv (e.g., in the Amazon forest, Posev 1985). Thus, in order to maintain this habitat diversity, nature conservation alone is not sufficient. Strategies must include a consensus on the objectives of conservation in a dynamic process as well as use of appro- priate conservation methods, including human action. Ultimately, the "integrity of nature," as claimed by biocentric holism, would change manmade landscapes into a less diverse pattern. In Figure 3 it was assumed that extinct species are unsubsti- tutable ("hard" sustainability), leading to time horizons on the order of magnitude of evolutionary processes. If substitutability of the value of species were assumed ("soft" sustainability), the temporal scope for their replacement would depend on the creativity of man (i.e., on technical progress). 40 The difference in temporal scope and speed between biological and technical processes is increasingly being recognized in the sustain- ability debate (Gowdy and Daniel 1995). The underlying hypothesis is that technical progress occurs much too quickly to be adjusted to bio- logical cycles, which ultimately govern all life and development on earth. Bund and Misereor (1996) thus concluded that it was necessary to demand "de-acceleration" (Entschleunigung) of development (cf. Gronemeyer 1993). Criteria for Indicator Selection Table 2 presents criteria for the selection and evaluation of sus- tainability indicators. The first demand on sustainability indicators is their scientific validity (BML 1995). Bernstein (1992) demanded that "the ideal trend indicator should be both ecologically realistic and meaningful and managerially useful." These two key properties should be complemented by the requirement that appropriate incdi- cators be based on the sustainability paradigm (cf. RSU 1994). This last property explicitly introduces the normative element, guiding selection of the indicator according to the value system of the respec- tive author, institution, or society. The requirements for sustainability indicators cover a broad range of aspects, not all of which can be equally met. This is most obvious in the category of ecosystem relevance (second column of Table 2). The first set of criteria in the second column (Ecosystem Relevance) describe desirable properties to determine the sustainability of ecosystems. However, they are not explicit with regard to system theory; thus, they still require operational- ization. The second group of indicators in column 2 is based on system theory; however, they cannot necessarily be put into prac- tice (e.g., large scale and long-term effects, such as global warrn- ing, are beyond the scope of experimental evidence from full sys- tem cycles). Thus, a balance must be found between scientific accuracy and pragmatic decisionmaking. Table 2 can be used to evaluate existing indicators by applying a matrix approach. Selected criteria of all columns can be matched with the indica- tors to be evaluated by assigning yes / no / not valid scores to each 41 'fable 2. Criteria for the Selection of Sustainability Indicators SCIENTIFIC QUALITY ECOSYSTEM RELEVANCE DATA MANAGEMENT SUSFAINABILI 1Y PARADIGM Indicator really mea- * Changes as the system * Easy to measure * What is to be sus- sures what it is sup- moves away from eqoii- * Easy to document tained? posed to detect librihLm * Easy to interpret * Resource efficiency * Indicator measures sig- * Distinguislses agroe- * Cost effective Carrying capacity titficanit aspect cosystelnTs ITovinlg * Data available * Health protection * Problern specific toward suistainiability * Comparable across bor- * Target values * Distiiguislses between * Idcnitifies kcy factors ders and over time * Time horizon causes and effects Icadinig to unsustaini- * Quiatitifiable * Social welfare * Can bc reproduced anid ability * Representative * Equlity repeated ovcr time * Warniing of irreversible * Iransparerst Participatory definition Uncorrelated, indepen- degradation processes * Geographically relevanit * Adequiate rating of sin- dent * Proactive in forecasting * Relevant to users gle aspects * Unambiguous future trends * User fricndly * Covers fuol cycle of the * Widely accepted system through time * Corresponds to aggrega- tion level * Highlights linkes to other system levels * Permits tradeoff detee- riots and assessment hetweeni SySierTi COMPO- nentts and levels * Cais be relatcd to octier inidicator s Sources: Bernstein (1992); BML (I 995); HFambliis (1992); 1 larriogtoiots et nl. (] 993); Mcissner- (1 993); Mille (1 995); Pcttit and Sar-wal (1993); RSU (1994); Via1 Kculeil (1993). (cf. RSU 1994). This can help guide the selection of indicators according to the purpose of the study. In Figure 3 most methodologies important to sustainability research are related to the spatial scale. From an ecological perspec- tive it can be concluded that biodiversity research, application of geo- graphic information systems (GIS) and remote sensing for determin- ing spatial characteristics, and modeling to assess the temporal dimension will become the dominant fields of sustainability researchi. Small-scale models will likely cover full svstem cycles with more or less well-researched determinants for temporal extrapolation (e.g., crop growth models), while large-scale models will in most cases cover long time spans, and therefore will need to incorporate risk assessment based on probability assumptions (cf. Ludwig et a!. 1993; Wissel 1995; Gowdy and Daniel 1995). From a social science point of view, Hallu and Runge-Metzger (1993) identified the following methodologies, but they did not relate them to spatial dimensions: (1) expert guesses (inter alia to capture uncertainty): Delphi techniques, consultations, and intelli- gent guesses (Meissner 1993 complemented this category with heuristic approaches); (2) experiments; (3) field surveys: social sci- ence, aerial surveys and remote sensing, direct field observations; and (4) modeling: empirical models, deterministic-analytical models, and deterministic-numerical models. From Sustainability Assessment to Sustainabifity Policy Ideally, policy decisions should coincide with the results of scien- tific analysis; however, because of conflicting interests this is frequently not the case. On the one hand, policy decisions are embedded in a cul- tural environment that is shaped by rational and by irrational tradi- tions, by ethical consensus and by discourse (Figure 1). In a modern society, on the other hand, policy decisions are based on scientific analysis, apart from the normative aspects (Figure 2). Thus, relating 43 scientific sustainability assessment to the cultural, ethical, and political context, it becomes apparent that scientific sustainability assessment can contribute significantly to policy decisionmaking, but that it is not the only basis for such decisions. Scientific analysis of sustainability and the selection of valid indi- cators serve as tools for rational decisionmaking and evaluation. Figure 3 shows the three major "clients" for sustainability indicators in the policy domain: farmers at the local level, industry managers at the local level to assess the environment of their plants and the mar- kets at the regional to global levels, and public policymakers (Hamblin 1992; Harrington et al 1993). They all need sustainabiliiy indicators for effective decisionmaking in space and time: between "here" and "there" and between "now" and "later" (Costanza 199 1; Redclift 1994). Dovers (1995) presented a general framework for scaling and framing policy problems in sustainability. He proposed a heuristic list of attributes for problem-framing and response-framing. Problem-framing is based on scientific sustainability assessment, while response-framing in the first place comprises normative and political criteria. Apart from criteria linked to spatial and temporal impacts, Dovers (1995) included the degree of complexity and connectivitv as a criteri- on for problem-framing. Similarly, the RSU (1994) pointed to the "retinity principle" (i.e., the interaction of components and dimen- sions in space and time) as the basic principle for sustainability, inte- grating economic, social, and ecological development. At the same time, the RSU admitted that the relationship between economy and ecology is inherently conflictive. As a result, this requires a consensus on values, including minimizing, but tolerating, (unavoidable) "evi."r and restrictions on current consumption. Forging conflicting interests into a coherent agenda based on max- imum consensus is the primary goal of democratic policy. Conflicts may arise between economy and ecology, between short-term and 44 long-term goals, and between local and global goals. These conflicts also can occur concurrently (connectivity, retinity). Specific policy recommendations and choices of specific policy measures are not the intent of this paper. Rather, its purpose is to pre- sent the basic values for sustainability assessment and to describe the set of instruments available for sustainability assessment, as well as the limitations of their application. Thus, this paper is restricted to prob- lem-framing; it does not address response-farming in detail (Dovers 1995). Ecology and Economy Because natural resources are scarce (Daly 1991), their monetary valuation has a significant, although limited, role in the policy domain. Monetary valuation, as a policy instrument, is necessary for resource allocation. In particular, it permits the internalization of externalities, and contributes to problem analysis from a cost-benefit point of view. With regard to the value of nature, anthropocentric principles may lead to valuations different from those of biocentric holism. For example, this may occur when a decision must be made about thie interests of the inhabitants of a mountain reserve versus the protec- tion of the gorillas living there. Careful analysis of such a situatioln, however, frequently reveals that respecting the rights of the human inhabitants also serves the creatures to be protected (cf. Hampicke 1994; Steger 1995). Three ethical principles for decisionmaking are proposed from an economic point of view: efficiency, sufficiency, and consistency. Efficiency is the most conventional policy recommendation. For example, the CGIAR Task Force on Sustainable Agriculture recom- mended that research focus on productivity and efficiency (CGIAR 1995). Proponents of this policy presume that technical progress will ultimately lead to such an efficient use of natural resources that depletion and pollution rates will no longer exceed the supply and absorption capacity of the environment. This is closely related to 45 "soft" sustainability, presuming the substitutability of natural resources by human intervention. On the contrary, sufficiency rec- ognizes the scarcity of unsubstitutable resources ("hard" sustainabili- ty). Sufficiency is the most controversial policy principle because it is associated with uncomfortable restrictions on wealth. The RSU (1994) demands working toward moral consensus in society to achieve long-term goals that are to be valued more than short-term wealth, although a combination, with restrictions and prohibitive measures, may be unavoidable. Consistency, related to "sensible" sustainability, has been proposed as a "third way," as economic management consistent with natural cvcles and phases (Huber 1995). It may, however, only pretend to harmonize economy and ecology. Ultimately, it consists of efficiency increase (though ideally cyclical) up to the limits of pool size and turnover rates when resource restrictions need to be faced (sufficiency). Specific policy measures will have to consider all three principles, or strategies. Efficiency increase may be applied in cases of lcw resource use. When applicable, consistency is a desirable strategy, but it is limited to cyclical processes. Ultimately, there is no alternative for restricting resource use (i.e., sufficiency) of limited, unsubsr- tutable natural resources. System theory for sustainability assessment can help identify resource pools, turnover rates, component interac- tions, externalities, and such. It also can shed light on the complexity and connectivity of policy problems, and on their hierarchical order. It cannot, however, replace political value judgment on the limits of use and policy strategies to achieve the envisaged goals. Short-Term and Long-Term Goals Translating the principle of intergenerational equity into policy requires one to define the priorities between short-term and long-term goals, to choose realistic time horizons, and to deal with uncertainty. Since the introduction of Nachhaltigkeit into forestry 200 years ago, the principle of present renunciation of wood harvest for the sake of securing future timber production has been claimed. This principle was forced on traditional smallholders by aristocratic landowne.rs, 46 depriving them of their source of fuel wood and fodder (Trossbach 1996). This situation is similar to that in which some contemporary policymakers of industrial countries demand restrictions on rain forest use by traditional forest dwellers. With regard to the temporal dimension of problem-framing, Dovers (1995) distinguished among the timing of possible impacts, their longevity, their reversibility, and their mensurability (i.e., their degree of risk, uncertainty, and ignorance). In modern society policy decisions generally have short time horizons (e.g., they are guided by interests in winning elections); that is, on the order of years rather than decades. Yet the impact of such decisions may have much longer effects; for example, in the case of nuclear energy (Meadows et al. 1992). Today's decisions frequently do not take into account the well-known potential for future damage. In the case of long-term environmental damage, Ludwig and others (1993) pointed out that "scientific certainty and consensus in itself would not prevent overexploitation and destruction of resources. Many practices continue even in cases where there is abundant sci- entific evidence that they are ultimately destructive." Using irriga- tion in the Euphrates and Tigris valley as an example, they con- cluded that "3000 years of experience and a good scientific under- standing of the phenomena, their causes, and the appropriate pro- phylactic measures are not sufficient to prevent the misuse and consequent destruction of resources." If policy decisions even in the case of well-known risk are not guided by advocating the interests of future generations, then it is even more difficult to institute preventive policies in cases of unknown risk, uncertainty, and ignorance of long-term impacts. For such cases the "precautionary principle" has been proposed. Dovers (1995) discussed the value of this principle for sustainability policies. He showed that- as has been shown for the ecosystem health approach-the precaution- ary principle is as normative and value-laden as sustainability itself, and is composed of loose, qualitative descriptors that are difficult to opera- tionalize. Proposed techniques for operationalization of the precaution- ary principle are limited in scale and scope (e.g., quantitative or ecolog- 47 ical risk assessment, minimax criteria, safe minimum standards, or environmental bonds [Dovers 19951). Dovers concluded that the pre- cautionary principle is "primarily a moral or political notion that may be informed (or misinformed!) by science." The decision on when to apply the principle and the definition of threshold values may be sup- ported by science, but it is still a normative, political decision based on informed judgment (cf. Wissel 1995). Local and Global Goals The majority of policy papers on global issues have their roots in industrial countries, while local strategies are proposed for agro- ecosystem management mainly in developing countries. This reflects the dominating interests and concerns: large-scale resource exploita- tion and pollution by industry versus agricultural production. Large-scale policy proposals tend to be closer to static, top-down (neoclassical) concepts, proposing target values for resource use, whereby target definition is strongly normative. For example, Levi (1995) recommended that estimates of future nuclear energy use be based on the assumption that the population of developing countries will reach a target energy consumption of one-quarter that of the current consumption levels by inhabitants of industrial nations. The opposite policy approach involves local agendas with broad participa- tion based on process-oriented, discursive development (RedcliEt 1994; Busch-L-ty 1995). Bottom-up policy strategies derived from local agendas have their limitations in a system hierarchy: many problems cannot be solved by simple aggregation of subsystems or components. Sustainabilicy assess- ment based on system theory can be used to define hierarchy struc- tures, connectivity, and system boundaries. There are few examples of spatial cross-boundary impacts of sustainability policies. Reardon and Vosti (1992) and Reardon and others (1995) analyzed such policy impacts on household-level decisionmaking in developing countries, as well as the interrela- tionship between environmental protection and poverty alleviation 48 (Reardon and Vosti 1995). Other cross-boundary examples include policies for interest compensation between local and large-scale goals, which may be developed by participatory optimization processes (Muller 1993). Although human society and social policymaking are distinct from ecosystem behavior, one can conclude-in analogy to biodiver- sity-that the advantage of a dynamic and participatory local agenda is that it maintains or increases cultural diversity and complexity, thus enhancing system stability and resilience (cf. Norgaard 1989). Conclusions Ludwig and others (1993), after their analysis of sustainability as a guiding policy principle for deep sea fisheries as a large-scale systema with a long time horizon, came, in brief, to the following conclusions: "Include human motivation, shortsightedness and greed; act before sci- entific consensus is achieved; rely on scientists to recognize problems but not to remedy them; distrust claims of sustainability; confront uncertainty; consider a variety of possible strategies; favor actions that are robust to uncertainties; hedge; favor actions that are informative; probe and experiment; monitor results; update assessments and modify policy accordingly; and favor actions that are reversible." Their condusions are confirmed by the analyses and results of this article, which follow: Be honest. Since its inception 200 years ago the concept of sus- tainability has been applied in the spirit of industrial development. It remains in this tradition in that its main focus is the repair of actual and potential environmental damages caused by industrial develop- ment (internalization of externalities). Sustainability principles can be upheld only as long as they do not interfere with dominating eco- nomic interests. Frequently, arguments for environmental protection in the interest of future generations deny contemporaries in develop- ing countries the same rights to resource consumption as the citizens of industrial nations. 49 Be modest. Scientific analysis of sustainability is a necessary and useful tool for problem-framing, but its role in the derivation of specif- ic policies is limited. Dovers (1995), following Funtowitz and Ravetz (1991), recognized a gradient for response-framing from a micro to a macro level. Micro-problems can be solved by "applied science," meso- issues by "professional consultancy," while macro-issues require "post- normal science," reflecting a gradient of increasing system uncertaintv and increasing decision stakes. Consequently, with increasing scale, policy decisions become more and more normative and value-guided, possibly even irrational, rather than based on "hard" scientific facts. Sustainability assessment of intergenerational equity should be claimed with care. The majority of methodologies for sustainability assessment of agroecosystems do not go beyond a couple of decades; thus, they do not cover intergenerational time spans (Figure 3). Similarly, the time horizon of policy measures (not impacts) has the same order of magnitude. If longer time spans are considered, the scientific operationalization of intergenerational equity is increasingly insecure. From the perspective of system theory it implies long-term ex ante analysis of (eco)system development, which lacks experimen- tal foundation. The claim of intergenerational equity is an ethically well-support- ed moral appeal, but it is not a commonly agreed upon policy princi- ple (cf. Agarwal 1993). It lacks sufficient ground for consensus in soci- ety because future generations cannot retaliate against present injustice. History has proven that greed and self-interest generally gain priority over unselfish, altruistic care for a yet nonexistent population. Be clear. Sustainability assessment demands clarity in two aspects. First, it requires the explicit definition of objectives, time scales, and spatial dimensions when applying scientific operational- ization. Only then can methodologies be adjusted to the objective and scope of the problem. Second, the underlying norms and values need to be clarified. Figure 2 shows that sustainable development is a strongly normative concept that requires scientific operationalization. For scientific sustainability assessment, the holistic ecosystem health 50 concept appears to be an appealing approach. Similarly, the precau- tionary principle appears to be an adequate policy for sustainable development. However, it has been shown that both concepts are as normative and vague as sustainability itself. Thus, scientific opera- tionalization of sustainability on the basis of either of these concepts needs special care. If applied with the necessary caution and with an awareness of their limitations, however, these normative concepts do have some justification. First, they are useful tools for defining targets and visions for policy development. (It should be made explicit, however, which normative decisions and values have entered into the definition of the target system.) Second, they show that reductionist science-even most complex modeling and system theory-cannot capture the full range of aspects associated with sustainability, just as a description of symptoms does not fully define health. Be cautious. Despite its operational limitations, the normative aspect of sustainability and its scientific assessment is a powerful and necessary concept for decisionmaking at all levels, from farming to international policy. Noting the potential long-term impacts of envi- ronmental management practices is certainly a prerequisite for deci- sionmaking, as long as alternative strategies exist and are viable. In view of potential damage to the environment, the precautionary prin- ciple is a useful tool despite its conceptual weakness. If possible, the reversibility of impacts should be aimed at. Careful scientific problem analysis using the presented methodologies for sustainability assess- ment can contribute to understanding often complex situations. Frequently, policy decisions are required without a full under- standing of the problem and its impact or without agreement on its scope. To limit the impact of policy decisions based on insufficient knowledge, discursive and iterative approaches are preferable. Participatory processes for seeking consensus on norms and strategies help define acceptable policy targets, although they do not guarantee that the diverse positions and interests of otherwise neglected groups will be considered. 51 References Adger, W. N., and F. Grohs. 1994. "Aggregate Estimate of Environmental Degradation for Zimbabwe: Does Sustainable National Income Ensure Sustainability?" Ecological Econonmics 11:93-104. Adriaanse, A. 1993. Environmental Policy Performance Indicators. A Study on the Development of Indicators for Environmental Policy in the Netherlands. The Hague: Uitgeverij. Agarwal, A. 1993. "The Best Solutions Are Home-Made." In Eco-Management-- The Hindu Survey of the Environment, edited by N. Ravi, 7-11. Madras: National Press. Barbier, E. B. 1989. Economics, Natural Resource Scarcity and Development: Conventional andAlternative Views. London: Earthscan Publication. Barbier, E. B., ed. 1993. Economics and Ecology. New Fronriers and Sustainable Development. London: Chapman and Hall. Barnett, V., R. Payne, and R. Steiner. 1995. Agricultural Sustainability. Economic, Environmental and Statistical Considerations. Chichester, U. K.: John Wilev and Sons. Becker, B. 1995. "Indicator Plants for Sustainability Assessment of Tropical Production Systems." Journal ofApplied Botany 69 (3/4): 145-151. Bernstein, B. B. 1992. "A Framework for Trend Detection: Coupling Ecological and Managerial Perspectives." In Ecological Indicators, edited by D. H1. McKenzie, D. E. Hyatt, and V. J. McDonald, 2 vols., 1101-1114. Proceedings of the International Svmposium on Ecological Indicators, October 16-19, 1990, Ft. Lauderdale, Florida. London and New York: Elsevier. Birnbacher, D., ed. 1980. Okologie undEthik. Stuttgart: Reclam. Birnbacher, D. 1989. "Okologie, Ethik und neues Handeln." In Pragmnatik, edit- ed by H. Stachowiak, vol. 3, 393-417. Hamburg: Felix Meiner Verlag. BML (Bundesministerium fur Ernahrung Landwirtschaft und Forsten). 1995. "Synoptic Portrait of Similarities of the Contents of Existing Criteria and Indicator Catalogues for Sustainable Forest Management." Backgrourd Paper by the Federal Ministry of Food, Agriculture, and Forestry, Bonn, pre- pared with assistance from the Federal Research Center for Forestry and Forest Industry. Boyle, T. 1995. "I1 situ Conservation Research and the CGIAR." CIFOR News 9:7. 52 Braun, J. V., and D. Virchov. 1995. "Okonomische Bewertung von Biotechnologie und Pflanzenvielfalt in Entwicklungslhndern." Entwicklung und 12ndlicher Raum 3:7-11. Brick, P. 1995. "Determined Opposition. The Wise Use Movement Challenges Environmentalists." Environment 37 (8): 17-20, 36-42. Brick, P. 1995. "The Wise Use Challenge: Reply to Commentaries." Environment 37 (9): 5, 43. Bund and Misereor, eds. 1996. Zukunftsfdhiges Deutschland. Ein Beitrag zu einer global nachhaltigen Entwicklung. Basel: Birkhauser. Busch-Luty, C. 1995. "Nachhaltige Entwicklung als Leitmodell einer okologis- chen Okonomie." In Nachbaltigkeit in naturwissenschaftlicher und sozialwis- senschafflicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 115- 126. Stuttgart: Hirzel. Carson, R. 1962. Silent Spring. Boston: Houghton Mifflin Company. CGIAR (Consultative Group on International Agricultural Research). 1995. "Report of the Task Force on Sustainable Agriculture." Document No. MTM/95/10 of the Mid-Term Meeting of the CGIAR, May 22-26, 1995, Nairobi, Kenya. Washington, D. C. Christen, 0. 1996. "Nachhaltige Landwirtschaft (Sustainable agriculture)-- Ideengeschichte, Inhalte und Konsequenzen fur Forschung, Lehre und Beratung." Berichte iiber Landwirtschaft 74:66-86. Conway, G. R. 1983. Agroecosystem Analysis. London: Imperial College of Science and Technology. . 1985. "Agroecosystem Analysis." AgriculturalAdministration 20:31-55. . 1993. "Sustainable Agriculture: The Trade-Offs with Productivity, Stability and Equitability." In Economics and Ecology. New Frontiers and Sustainable Development, edited by E. B. Barbier, 46-65. London: Chapman and Hall. Conway, G. R., and E. B. Barbier. 1990. After the Green Revolution. Sustainable Agriculturefor Development. London: Earthscan. Costanza, R., ed. 1991. Ecological Economics: The Science and Management of Sustainability. New York: Columbia University Press. Costanza, R. 1993. "Ecological Economic Systems Analysis: Order and Chaos." In Economics and Ecology. New Frontiers and Sustainable Development, edited by E. B. Barbier, 29-45. London: Chapman and Hall. Costanza, R., H. E. Dalv, and J. A. Bartholomew. 1991. "Goals, Agenda and Policy Recommendations for Ecological Economics." In Ecological Economics: The Science and Management of Sustainability, edited by R. Costanza, 1-21. New York: Columbia University Press. 53 Dalsgaard, J. P. T., C. Lightfoot, and V. Christensen. 1995. "Towards Quantification of Ecological Sustainability in Farming Systems Analysis." Ecological Engineering 4:181-189. Daly, H. E. 1991. "From Empty-World to Full-World Economics: Recognizing a Historical Turning Point in Economic Development." In Environmentally Sustainable Economic Development: Building on Brundtland, edited by R. Goodland, H. E. Daly, S. El Serafy, and B. von Droste. Paris: United Nations Educational Scientific and Cultural Organisation (UNESCO). De Boef, W., K. Amanor, K. Wellard, and A. Bebbington. 1993. Cultivating Knowledge. Genetic Diversity, Farmer Experimentation and Crop Research. London: Intermediate Technology Publications. Dietz, F. J., U. E. Simonis, and J. van der Straaten, eds. 1992. Sustainability a'zd Environmnental Policy. Restraints and Advances. Berlin: Edition Sigma. Dorner, D. 1989. Die Logik des Miflingens. Strategisches Denken in koinplexen Situationen. Reinbek: Rowohlt. Dovers, S. R. 1995. "A Framework for Scaling and Framing Policv Problems in Sustainabilitv." Ecological Economics 12:93-106. Ehui, S. K., and D. S. C. Spencer. 1990. Indices for Mveasuring the Sustainability and Economic Viability of Farming Systems. Resource and Crop Management Program (RCMP) Research Monograph No. 3. Ibadan: International Institute of Tropical Agriculture (IITA). Engelhardt, T., and H. Waibel. 1993. "Umwelt6konomie in der land- wirtschaftlichen Technischen Zusammenarbeit." Der Tropenlandwirt 94:133-143. Feyerabend, P. 1987. Farewell to Reason. London: Verso. Flitner, M. 1995. Sammler, RPuber und Gelehrte. Die politischen Interessen an pflanzengenetischen Ressourcen 1895-1995. Frankfurt and New York: Campus Verlag. Foucault, M. 1981. Arch20logie des Wissens. Frankfurt: Suhrkamp. Fraser-Darling, F. 1969. "Man's Responsibilitv for the Environment." In Biology and Ethics, edited by F. J. Ebling, 117-122. London: Academic Press. Fresco, L. 0. 1994. "Agro-Ecological Knowledge at Different Scales." In Eto- Regional Approaches for Sustainable Land Use and Food Production, edited by J. Bouma, A. Kuyvenhoven, B. A. M. Bouman, J. C. Luyten, and H. G. Zandstra, 133-142. Dordrecht, Boston, and London: Kluwer and Centro Internacional de la Papa (CIP). Fresco, L. O., and S. B. Kroonenberg. 1992. "Time and Spatial Scales in Ecological Sustainability." Land Use Policy 9:155-168. 54 Fritz, P., J. Huber, and H. W. Levi. 1995. "Das Konzept der nachhaltigen Entwicklung als neue Etappe der Suche nach einem umweltvertraglichen Entwicklungsmodell der modernen Gesellschaft." In Nachhaltigkeit in natur- wissenschaftlicher und sozialwissenschafflicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 7-16. Stuttgart: Hirzel. Funtowicz, S. O., and J. R. Ravetz. 1991. "A New Scientific Methodology for Global Environmental Issues." In Ecological Economics: The Science and Management of Sustainability, edited by R. Costanza, 137-152. New York: Columbia University Press. Gaston, K. J., ed. 1996. Biodiversity. A Biology of Numbers and Difference. Oxford: Blackwell Science. Gowdy, J. M., and C. N. McDaniel. 1995. "One World, One Experiment: Addressing the Biodiversity-Economics Conflict." Ecological Economics 15:181-192. Graham-Tomasi, T. 1991. "Sustainability: Concepts and Implications for Agricultural Research Policy." In Agricultural Research Policy. internationial Quantitative Perspectives, edited by P. G. Pardey, J. Roseboom, and J. R. Anderson, 8 1-102. Cambridge: International Service for National Agricultural Research (ISNAR) and Cambridge University Press. Gronemeyer, M. 1993. Das Leben als letzte Gelegenheit. Darmstadt: Primus. Haber, W. 1994. "Ist 'Nachhaltigkeit' (sustainabilit,y) ein tragfihiges okologisches Konzept?" Verhandlungen der Gesellschafffur Okologie 23:7-17. Haber, W. 1995. "Das Nachhaltigkeitsprinzip als 6kologisches Konzept." In ANachhaltigkeit in naturwissenschafilicher und sozialwissenischafrlicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 17-30. Stuttgart: Hirzel. Hailu, Z., and A. Runge-Metzger. 1993. Sustainability of Land Use Systems. The Potential of Indigenous Measures for the Maintenance of Soil Productivity in Sub-Saharan African Agriculture. Weikersheim: Margraf. Hamblin, A., ed. 1992. "Environmental Indicators for Sustainable Agriculture." Report on a National Workshop, November 28-29, 1991. Bureau of Rural Resources, Land and Water Resource Research and Development Corporation, Grains Research and Development Corporation, Canberra, Australia. Hammond, A., A. Adriaanse, E. Rodenburg, D. Bryant, and R. Woodward. 1995. Environmental Indicators: A Systematic Approach to Measuring and Reporting on Environmental Policy Performance in the Context of Sustainable Development. Washington, D.C.: World Resources Institute. Hampicke, U. 1993. "Naturschutz und Ethik-Ruckblick auf eine 20jahrige Diskussion, 1973-1993, und politische Folgerungen." Zeitschri fur Okologie undNaturschutz 2:73-86. 55 Hampicke, U. 1994. "Ethics and Economics of Conservation.' Biological Conservation 67:219-231. Hardin, G. 1991. "Paramount Positions in Ecological Economics." In Ecological Economics: The Science and Management of Sustainability, edited by R. Costanza, 47-57. New York: Columbia University Press. Harrington, L., P. G. Jones, and M. Winograd. 1993. "Measurements and Indicators of Sustainability. Report of a Consultancy Team." Cent-o Internacional de Agricultura Tropical (CIAT), Cali, Colombia. Heinen, J. T. 1994. "Emerging, Diverging and Converging Paradigms cn Sustainable Development." International Journal of Sustainable Detvelopment and World Ecology 1 (1): 22-33. Henseling, K. O., and E. Schwanhold. 1995. "Eine nachhaltige zukun- ftsvertragliche Stoffwirtschaft als politisches Leitbild." In Nachhaltigkeit in naturwissenschaftlicher und sozialwissenschafrlicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 81-90. Stuttgart: Hirzel. Herdt, R. W.. and J. K. Lynam. 1992. "Sustainable Development and tl.e Changing Needs of International Agricultural Research." In Assessing the Impact of International Agricultural Research for Sustainable Developmer t, edited by D. R. Lee, S. Kearl, and N. Uphoff. Symposium Proceedings, Cornell University, June 16-19, 1991, Ithaca, New York. Cornell Institute for Food, Agriculture, and Development. Holling, C. S. 1987. "The Resilience of Terrestrial Ecosystems: Local Surprise and Global Change." In Sustainable Development of the Biosphere, edited by W. C. Clark and R. E. Munn, 292-317. London: Cambridge University Press. Huber, J. 1995. "Nachhaltige Entwicklung durch Suffizienz, Effizienz und Konsistenz." In Nachhaltigkeit in naturwissenschafilicher und sozialwissenschafrlicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 31-46. Stuttgart: Hirzel. IBSRA.M (International Board for Soil Research and Management). 1994. Soil, Water and Nutrient Management Research-A New Agenda. IBSRAM Position Paper No. 1. Bangkok. IIMI (International Irrigation Management Institute). 1995. "A Proposal to the Technical Advisory Committee from the Inter-Center Initiative on Water Management.' Document P3 of Agenda Item 6 of the Twentieth Meeting of the Board of Governors of IIMI, December 6-8, 1995, Lahore, Pakistan. Colombo, Sri Lanka. Ikerd, J. E. 1993. "The Need for a Svstems Approach to Sustainable Agriculture." Agriculture, Ecosystems and Envi;-onment 46:147-160. Izac, A.-M. N., and M. J. Swift. 1994. "On Agricultural Sustainability and Its Measurement in Small-Scale Farming in Sub-Saharan Africa." Ecological Economics 11:105-125. 56 Jonas, H. 1984. Das Prinzip Verantwortung-Versuch einer Ethikfuir die technologis- che Zivilisation. Frankfurt am Main: Insel Verlag. Also published in 1984 under the title The Imperative of Responsibility, translated by Hans Jonas, with the collaboration of David Herr. Chicago: University of Chicago Press. Kant, I. [1785] 1961. Grundlegung ter Metaphysik der Sitten. (Fundamental prin- ciples of the metaphysics of morals). Stuttgart: Reclam. Kuhn, T. S. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press. -_____ .1969. The Structure of Scientific Revolutions. International Encyclopedia of Unified Science. Chicago: University of Chicago Press. Kuik, O., and H. Verbruggen, eds. 1991. In Search of Indicators of Sustainable Development. Environment and Management No. 1. Dordrecht, the Netherlands: Kluwer. Lantermann, E.-D. 1996. "Nachhaltigkeit als Leitlinie interdisziplinsrei Umweltforschung." Paper presented at a seminar on "Nachhaltige Ressourcennutzung," July 11, 1996, Kassel University, Witzenhausen, Germany. Lele, S. M. 1991. "Sustainable Development: A Critical Review." World Development 19:607-621. Leopold, A. [1949] 1989. A Sand County Almanac and Sketches Here and There. Special commemorative edition. Oxford, New York: Oxford University Press. Levi, H. W. 1995. "Das Problem der Nachhaltigkeit in der Energieversorgung." In Nachhaltigkeit in naturwissenschaftlicher und sozialwissenschaftlicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 47-58. Stuttgart: Hirzel. Ludwig, D., R. Hilborn, and C. Walters. 1993. "Uncertainty, Resource Exploitation, and Conservation: Lessons from Historv." Science 260:17, 36. Lynam, J. K., and R. W. Herdt. 1989. "Sense and Sustainability: Sustainability as an Objective in International Agricultural Research." Agricultural Economics 3 (4): 381-398. McKenzie, D. H., D. E. Hyatt, and V. J. McDonald, eds. 1992. Ecological Indicators. 2 vols. Proceedings of the International Symposium on Ecological Indicators, October 16-19, 1990, Ft. Lauderdale, Florida. London and New York: Elsevier. Meadows, D. H., D. L. Meadows, J. Randers, and W. W. Behrens III. 1972. The Limits to Growth. New York: Universe Books. Meadows, D. H., D. L. Meadows, and J. Randers. 1992. Beyond the Limits. Postmill, Vermont: Chelsea Green Publishing Company. 57 Meissner, M. 1993. "Beurteilungsmoglichkeiten von Nachhaltigkeit. Zur la sd- wirtschaftlichen Relevanz er-weiterter Meg- und Informationssysteme als Beurteilungskriterien der Nachhaltigkeit." Diplomarbeit, Gesamthochschule Kassel, Witzenhausen, Germany. Maller, H. P. 1993. "Okologische Ethik und Umweltrecht: Beginn und Umsetzung einer gesellschaftlichen Neuorientierung?" Paper presented at seminar on "Umweltrecht," July 1993, Gesamthochschule Kassel, Witzenhausen, Germany. Muiller, S. 1995. "Evaluating the Sustainability of Agriculture at Different Hierarchical Levels: A Framework for the Definition of Indicators." Working Paper. Instituto Interamericano de Cooperaci6n para la Agricultura (IICA), San Jose, Costa Rica. Nelson, R. H. 1995. "Sustainability, Efficiency, and God: Economic Values aLnd the Sustainability Debate." Annual Review ofEcology and Systematics 26:1.35- 154. Norgaard, R. B. 1989. "The Case for Methodological Pluralism." Ecolog:cal Economics 1:37-57. Norton, B. G. 1991. "Ecological Health and Sustainable Resource Management." In Ecological Economics: The Science and Management of Sustainability, edited by R. Costanza, 102-117. New York: Columbia University Press. OECD (Organisation for Economic Co-operation and Development). 1991. Environmental Indicators. A Preliminaiy Set. Paris. Olembo, R. 1994. "Can Land Use Planning Contribute to Sustainabilityi" In The Future of the Land: Mobilising and Integrating Knowledge for Land Use Options, edited bv L. 0. Fresco, L. Stroosnijder, J. Bouma, and H. Van Keulen, 369-376. Chichester, U.K.: John Wiley and Sons. Pandev, S., and J. B. Hardaker. 1995. "The Role of Modelling in the Quest for Sustainable Farming Systems." Agricultural Systcenis 47:439-450. Pardey, P. G., J. Roseboom, and 1. R. Anderson, eds. 1991. Agricultural Research Policy. International Quantitative Perspectives. Cambridge, U.K.: International Service for National Agricultural Research (ISNAR) and Cambridge University Press. Pareto, W. 1935. The Mind and Society. New York: Harcourt. Pearce, D. W.. E. B. Barbier, and A. Markandya. 1990. Sustainable Development: Economics and Environment in the Third World. Aldershot, U. K.: Edward Elgar. Perrings, C. 1991. "Reserved Rationalitv and the Precautionary Principle: Technological Change, Time and Uncertainty in Environmental Decision Making." In Ecological Economics: The Science and Management of Sustainabiiity, edited by R Costanza, 153-167. New York: Columbia Universin' Press. 58 Peters, W., and C. Wiebecke. 1983. "Die Nachhaltigkeit als Grundsatz der Forstwirtschaft." Forstarchiv 54 (5): 172-178. Pettit, B., and T. Sarwal. 1993. "Identifying Indicators of Sustainability: A Methodology." Draft paper. School of Development Studies, University of East Anglia, United Kingdom. Piepho, H.-P. 1996. "Einige Oberlegungen zur Quantifizierung von Nachhaltigkeit unter besonderer Beriicksichtigung von Biodiversitat und Stabilitat. In Nachhaltige Ressourcennutzung, edited by P. Wolff, 137-162. Beiheft Nr. 56 zum Tropenlandwirt. Witzenhausen: Selbstverlag des Verbandes der Tropenlandwirre Witzenhausen. Pieri, C., J. Dumanski, A. Hamblin, and A. Young. 1995. Land Quality Indicators. World Bank Discussion Paper 315. Washington, D.C. Plucknett, D. L., and N. J. H. Smith. 1986. "Sustaining Agricultural Yields." BioScience 36 (1): 40-45. Posey, D. A. 1985. "Indigenous Management of Tropical Forest Ecosystems: The Case of the Kayapo Indians of the Brazilian Amazon." Agroforestry Systems 3: 139-157. Powell, T. M. 1995. "Physical and Biological Scales of Variability in Lakes, Estuaries, and the Coastal Ocean." In Ecological Time Series, edited by T. M. Powell and J. H. Steele, 119-138. New York: Chapman and Hall. Rapport, D. J. 1992. "Evolution of Indicators of Ecosystem Health." In Ecological Indicators, 2 vols., edited by D. H. McKenzie, D. E. Hyatt, and V. J. McDonald, 121-134. Proceedings of the International Symposium on Ecological Indicators, October 16-19, 1990, Ft. Lauderdale, Florida. London and New York: Elsevier. Rawls, J. 1971. A Theory of Justice. Cambridge, Massachusetts: Harvard University Press. Reardon, T., E. Crawford, V. Kellv, and B. Diagana. 1995. Promoting Farm Investmentfor Sstainable African Agriculture. MSU International Development Papers No. 18. East Lansing, Michigan: Michigan State University, Department of Agricultural Economics, Department of Economics. Reardon, T., and S. A. Vosti. 1992. "Issues in the Analysis of the Effects of Policy on Conservation and Productivity at the Household Level in Developing Countries." Quarterly Journal of InternationalAgriculture 31 (4): 380-396. Reardon, T., and S. A. Vosti. 1995. "Links between Rural Poverty and the Environment in Developing Countries: Asset Categories and Investment Poverty." World Development 23 (9). 59 Redclift, M. 1994. "Sustainable Development: Economics and the Environmer[t." In Strategies for Sustainable Development: Local Agendas for the Southern Hemisphere, edited by M. Redclift and C. Sage, 17-34. Chichester, U.K.: John Wiley and Sons. . 1994a. "Reflecrions on the 'Sustainable Development' Debate." Internationaljournal of Sustainable Development and World Ecology 1 (1): 3-21. Regier, H. A. 1992. "Indicators of Ecosystem Integrity." In Ecological Indicators, 2 vols., edited by D. H. McKenzie, D. E. Hyatt, and V. J. McDonald, 183-200. Proceedings of the International Symposium on Ecological Indicators, October 16-19, 1990, Ft. Lauderdale, Florida. London and New York: Elsevier. Reid, W., J. McNeely, D. Tunstall, D. Bryant, and M. Winograd. 1992. Developing Indicators of Biodiversity Conservation. Washington, D.C.: World Resources Institute. Rbling, N. 1994. "Platforms for Decision-Making About Ecosystems." In The Future of the Land: Mobilising and Integrating Knowledge for Land Use Options, edited by L. 0. Fresco, L. Stroosnijder, J. Bouma, and H. Van Keulen, 385-393. Chichester, U.K.: John Wiley and Sons. RSU (Der Rat von Sachverstandigen fur Umweltfragen). 1994. Umweltgutachten 1994. Stuttgart: Verlag Metzler-Poesche. Ruttan, V. 1994. "Constraints on the Design of Sustainable Systems of Agricultural Production." Ecological Economics 10:209-219. Sachs, W. 1993. "Die Kunst des Unterlassens ist eine Tugend." Der Uberblick 4:95-97. Schonwiese, C.-D. 1995. "Nachhaltige Entwicklung des Klimas." In Nachhaltigkeit in naturwissenschaftlicher und sozialwissenschaftlicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 59-68. Stuttgart: Hirzel. Schulze, E.-D., and H. A. Mooney, eds. 1993. Biodiversity and Ecosystem Functirn. Ecological Studies 99. Berlin, Heidelberg, and New York: Springer Verlag. Schweitzer, A. 1984. Die Ehrfurcht vor dem Leben. Grundrexte aus fiinf Jahrzehnten, edited by H. W. Bahr. Munchen: Beck. Shearman, R. 1990. "The Meaning and Ethics of Sustainability." Environmental Management 14 (1): 1-8. Shrader-Frechette, K. S. 1994. "Ecosystem Health: A New Paradigm for Ecological Assessment?" Trends in Ecology and Evolution 9 (12): 456-457. Spendjian, G. 1991. "Economic, Social, and Policy Aspects of Sustainable Land Use." In Evaluation for Sustainable Land Management in the Developing World. Vol. 2, Technical Papers, 415-436. Proceedings No. 12. Bangkok: International Board for Soil Research and Management (IBSRAM). 60 Steger, U. 1995. "Nachhaltige und dauerhafte Entwicklung aus wirtschaftswis- senschaftlicher Sicht." In Nachhaltigkeit in naturwissenschaftlicher und sozial-- wissenschaftlicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 91-98. Stuttgart: Hirzel. Stoippler-Zimmer, H. 1993. "Ober die Schwierigkeiten, partizipative Technikfolgenabschatzung und -bewertung (TA) in die Praxis umzusetzen." GAIA 2 (4): 188-192. Swaminathan, M. S. 1991. "Sustainable Agricultural Systems and Food Security." Outlook on Agriculture 20 (4): 243-249. Tabora Jr., P. C. 1991. "Analysis and Evaluation of Agroforestry as an Alternative Environmental Design in the Philippines." Agroforestry Systems 14:39-63. Technical Advisory Committee, CGIAR (Consultative Group on International Agricultural Research). 1989. Sustainable Agricultural Production: Implications for International Agricultural Research. Food and Agriculture Organization of the United Nations (FAO) Research and Development Paper No. 4. Rome. ten Brink, B. 1991. "The AMOEBA Approach as a Useful Tool for Establishing Sustainable Development?" In In Search of Indicators of Sustainable Development, edited by 0. Kuik and H. Verbruggen, 71-88. Environment and Management No. 1. Dordrecht, the Netherlands: Kluwer. Tisdell, C. 1988. "Sustainable Development: Differing Perspectives of Ecologists and Economists, and Relevance to LDCs." World Development 16 (3): 373- 384. Trossbach, W. 1996. "Gelichtete Walder, verstiimmelte Eichen." In Nachhaltige Ressourcennutzung, edited by P. Wolff, 51-72. Beiheft Nr. 56 zum Tropenlandwirt. Witzenhausen: Selbstverlag des Verbandes der Tropenlandwirte Witzenhausen. Turner, R. K., and D. W. Pearce. 1993. "Sustainable Economic Development: Economic and Ethical Principles." In Economics and Ecology, New Frontiers and Sustainable Development, edited by E. B. Barbier, 177-1940. London: Chapman and Hall. Van Keulen, H. 1993. "Long-Term Modelling and Monitoring of Sustainability." Presentation at an internal meeting of the Centro International de la Papa (CIP), October 13, 1993, Lima, Peru. van Pelt, M. J. F., A. Kuyvenhoven, and P. Nijkamp. 1995. "Environmental Sustainability: Issues of Definition and Measurement." InternationalJournal of Environment and Pollution 5 (2/3): 204-223. Vedeld, P. 0. 1994. "The Environment and Interdisciplinarity. Ecological and Neoclassical Approaches to the Use of Natural Resources." Ecological Economics 10:1-13. 61 Walsh, J. 1991. Preserving the Options: Food Productivity and Susrainability. Issues in Agriculture No. 2. Washington, D.C.: Consultative Group on International Agricultural Research (CGIAR). WCED (World Commission on Environment and Development). 1987. Our Comnmon Future. The Brundrland Report. Oxford, U.K.: Oxford University Press. Wiersum, K. F. 1995. "200 years of Sustainability in Forestry: Lessons from History." Environmental Management 19 (3): 321-329. Wilson, E. O., ed. 1988. Biodiversity. Washington, D.C.: National Academv Press. Winograd, M. 1995. Environmental Indicators for Latin America and the Caribbean: Towards Land-Use Sustainability. GASE (Ecological Systems Analysis Group, Bariloche, Argentina), in collaboration with IICA (Instituto Interamerican de Cooperacion para la Agricultura)-GTZ (Deutsche Gesellschaft fur Technische Zusammenarbeit) Project, OAS (Organisation of American States), and WRI (World Resources Institute). Washington, D.C. Wissel, C. 1995. "Nachhaltigk-eit aus der Sicht der okologischen Mlodellierung." In Nachhaltigkeir in naturwissenschaftlicher und sozialwissenschaftlicher Perspektive, edited by P. Fritz, J. Huber, and H. W. Levi, 127-132. Stuttgart: Hirzel. Wolff, P. 1994. "Zum Problem der Nachhaltigkeit von Bewasserungssystemen." In N7Vachhaltigkeit der Entwickluna ldndlicher Regionen Afirikas, Asiens ,Ind Lateinamerikas, edited by S. Amini and E. Baum, 57-67. Beiheft 52 zum Tropenlandwirt. Witzenhausen: Selbstverlag des Verbandes der Tropenlandwirte Witzenhausen. Wolters, J. 1995. "Die Arche wird gepliindert. Vom drohenden Ende der biolo- gischen Vielfalt." In Leben und leben lassen: Biodiversitat - Okonomie, Natur- und Kulturschutz im Widerstreit, edited by J. Wolters, 11-39. Okozid No. 10. Giessen: Focus. 62 About the Author Barbara Becker is a Senior Scientist at the Institute for Crop Science of the University of Kassel, Federal Republic of Germany. Her major field of research is agroecology, in particular vegetation analysis of the Andean highlands related to agricultural production systems. She gained experience on high Andean agroecology while serving as a UNEP Field Project Officer in Peru between 1985 andi 1989. Her present work on Andean agroecology is carried out in the framework of CONDESAN, in collaboration with CIP and CIAT. Ms. Becker's education included training in mathematics, botany, and tropical agronomy at Goettingen University, where she received her Ph.D. for research on wild plants for human nutrition in. the arid zones of Sub-Saharan Africa (Kenya and Senegal). During that time she acquired experience with agroforestry, which is one of her current teaching subjects. She is a member of the Editorial Board of "Agroforestry Systems." Before joining Kassel University, Ms. Becker was Liaison Officer for International Agricultural Research at the German Council for Tropical and Subtropical Agricultural Research (ATSAF e.V.). She became familiar with the CGIAR system as a. member of the German donor delegation and as an observer at Technical Advisory Committee (TAC) meetings. She was recently appointed a member of the Board of Trustees of the International Board for Soil Research and Management (IBSRAM) in Thailand. 63